School absence: Bums-in-seats-thinking is the wrong problem, and AI the wrong solution

What school absence trends do you believe AI can spot that current methods cannot, to improve coordination between education, social care and the wider services that support families?

Since the pandemic, conservative leaning organisations and most notably the CSJ, have been banging a drum very loudly about children not in school. This is not about private-school educated children, but children they expect be in state educational settings age 5-18. Often, other people’s children.

In recent debate, various sub-categories of children not in school are conflated together. In 2022, various stakeholders got together and with Defend Digital Me, we worked to point out in non-partisan way, where that thinking was wrong.

It is wrong on who is not counted. (And who is). It is wrong on assumptions built into conflating the differences between children not in school and children ‘not in receipt of suitable education’.  It is wrong on assumptions that children not in school are not already on registers and recorded on Local Authority databases. It is wrong to ignore that the definition of persistent absence means children are classified as persistently absent more quickly now than previously. The same number of children could be absent for the same number of school days as were a decade ago, but it will be reported as having doubled.

And it is arguably morally wrong, in ignoring a very significant part of the problem as they see it: why children are not in school once you remove all that is wrong with the counting and assumptions. And what are the consequences for children of forcing them to be without their consent, or respecting  families choice or agency?

Absence data (on school roll, not attending)

In the 2023/24 academic year up to 8 December 2023, DfE data shows that the attendance rate across the academic year to date was 93.4%. The absence rate was, therefore, 6.6% across all schools and the unauthorised rate was far less. By school type, the absence rates across the academic year 2023/24 to date were:

  • 5.1% in state-funded primary schools (3.7% authorised and 1.4% unauthorised)
  • 8.3% in state-funded secondary schools (5.2% authorised and 3.1% unauthorised)
  • 12.6% in state-funded special schools (9.5% authorised and 3.0% unauthorised)

Over 1.5 million pupils in England have special educational needs (SEN)

An increase of 87,000 from 2022. Both the number of pupils with an education, health and care (EHC plan) and the number of pupils [recorded] with SEN support have increased:

  • The percentage of pupils with an EHC plan has increased to 4.3%, from 4.0% in 2022.
  • The percentage of pupils with SEN but no EHC plan (SEN support) has increased to 13.0%, from 12.6% in 2022.

Both continue a trend of increases since 2016. As do the number of stories you hear of parents asked to bring in children only part-time because the schools cannot get EHC plans approved, (Local Councils have no money) without which, schools cannot access funds or allocate the staff and resources needed for that child in a school.

Which children are not in school?

The concept of children-not-in-school should be nothing at all to do with Elective Home Education (“EHE”). The premise of ‘not in school’ is that they are not attending school ‘but they should be’ and that action taken will be as a result. Elective Home Education (“EHE”) children are not on a school roll and not expected to be. No action should be taken as a result, to get them into schools.

A Guardian article today (Jan 9th) quotes Wendy Charles-Warner, chair of home education charity Education Otherwise, who sums up one problem here: “Yet again we see an inappropriate and frankly mangled conflation of [elective] home education and absenteeism.

“Home education is of equal legal status to school education and it is certainly not ‘non-attendance’. Home educated children are in full-time education, they are not school pupils let alone absent school pupils.

“A register of home-educated children will make no difference whatsoever to school absenteeism and, before proposing such a significant step, the Labour party should educate itself to the very basic facts of the matter.”

It was frustrating to hear Bridget Phillipson give the same impression today as many other MPs had in the 2022 debates on the Schools Bill but using selective evidence, that no one knows how many children are home educated. Every Local Authority we asked data for in 2021-22 that replied, already had a register of EHE.

Proposals to legislate for a new national register of children not in school were part of the Government’s now-scrapped Schools Bill and were hotly contested and debated in the House of Lords. There is no compelling case to have one.

We assessed the plans at Defend Digital Me as part of the Counting Children coalition, and not only were the policy issues pretty fundamentally flawed, but practically flawed too. The legislation on the database as set out would have meant for example, double counting a whole swathe of children already on school registers but in Alternative Provision or part-time.

The plans conflated Home Education (“EHE”) (not on school roll) with absenteeism (pupils registered on school roll but absent), and would have doubled counted some children who were part time at alternative school settings, and conflated these children with Children Missing Education (“CME”, not on a school roll and it has been assessed that they are ‘not in receipt of a suitable education otherwise’, so should be on a school roll but are not and who are known to the State); and further conflated those three groups with children not on any database at all (unknown to the state education system).

Piling in elective home educators with at-home children waiting for places or suitable state school services, with children already on roll but part-time and truancy, would have geared up to conflate a toxic mix of ‘victims’ and ‘perpetrators’ style narrative like the Met Police Gangs matrix, people to be treated with suspicion and requiring additional (often centralised) state surveillance, all as a result of what would have been bad numbers.

The not-in-school register plans angered many, among others, the Charedi community. Home Educators protested outside parliament and filled the public gallery in the House of Lords on the day it was most specifically debated in the Schools Bill.

I know of cases where children are wrongly labeled CME by Local Authorities. Some LAs cannot (for what seems like nothing but stubborn jobsworth bureaucracy) accept that children who are Home Educated can be in receipt of suitable education, even if an LA’s opaque methods of measuring ‘suitability’ are so arbitrary and intrusive and out of step with the law as to be only understandable to them. Such records create fundamentally flawed and inaccurate family portraits, turning EHE records into CME records mid-year, (spot the double counted child) without recourse or route for redress. Some of these data are therefore opinions, not facts. Automating any of this would be a worse mess.

Known-to-the state CME children are on Local Authority databases even if If they were never enrolled at a school. If they were in school and left and even if after doing so they ‘fall off the radar’, one of sixteen tick-box reasons-for-leaving is recorded on their detailed named record and kept. If they disappear without a known destination of the next educational setting, they will in addition be added into the part of the Common Transfer File system that posts children’s records into the Lost Pupils Database (LPD). They are pulled out of that LPD once a state school ‘receives’ them again.

Children about whom nothing is known by the state, the so called ‘invisible children’ cannot be magically added to any database. If they were known today, they would already be on the existing databases. It is believed there are very few of these, but of course, it is unknown. It also by default a number that cannot ever be known. The NCB and Children’s Commissioner have made guesstimates of around 3,000 individual children.

The CSJ has in my view whether accidentally or by intent, wrongly hyped up the perception of the numbers of those children, by inventing the new term “ghost children”. This has made the everyday listener or MP think of these as ‘unseen by the state’. This CSJ term has been sweepingly used in the media and parliament to cover any child not in school and means the perceived “problem” is wrongly seen as (a) one and the same thing and (b) much larger than in reality.

That reshaping of this reality matters. It’s been a persistently retold half-truth since 2021 (the Telegraph published my letter to the editor on it in March 2022).  Still it seems not easily fixed by fact alone. The costs of new databases duplicating data that already exist, would be far better spent on patching up the 70% cuts to Local Authorities youth services, CAMHS, or Early Intervention Grant, or basically anything else for children and young people or families.

Which children do we know should be in school?

Remembering that the claims are that we need new registers of children not-in-school, how many children do you think are known to be missing education, recorded as CME, in any one Local Authority? Yes, these children are recorded by name already at LAs.

Local authorities have a duty under section 436A of the Education Act 1996 to make arrangements to identify, as far as it is possible to do so, children missing education. What is possible, is already done.

There are currently 152 local education authorities in England and through the dedicated volunteer effort from the Counting Children coalition, we asked all of them for their data (as of June 30, 2021).

There were zero Children Missing Education (“CME”) in Powys, Wales. In Blackpool by contrast lots. There were 45 Children Missing Education (“CME”) (in the area waiting for provision to start, mainly recently arrived), 112 Children missing “Out” (left area being tracked) (of which 61 had been located) and 307 Elective Home Education (“EHE”). And across the academic year September 2020 – July 2021, the Isle of Wight recorded 49 children as Children Missing Education (CME), in East Riding there were 17 Children Missing Education (“CME”). In Leicester they even noted that children on their registers have been recorded in these ways since 2003.

Do those numbers surprise you? Local Authorities also collect a lot of data already about each child out of school. For example, Harrow’s central database on children not in school already includes Family Name, Forename, Middle name, DOB, Unique Pupil Number (“UPN”), Former UPN (For adopted children and children-at-risk this should not be so, but who knows if it is respected see 6.5 and 6.6 due to risks the UPNs create for those children — The UPN is supposed to be secure state identifier, and as such, has special protections including being a blind identifier and it should lapse when children leave school (see page 6-8).  There is also the  Unique Learner Number (ULN).

Any new number policymakers suggest inventing, would need to be subject to the same protections for the child (and throughout their adult life), and therefore it would serve little purpose to create yet another new number.

The list goes on, of what is collected in CME Local Authority databases on each named child. Address (multi-field), Chosen surname, Chosen given name, NCY (year group), Gender, Ethnicity, Ethnicity source, Home Language, First Language, EAL (English as additional language), Religion, Medical flag, Connexions Assent, School name, School start date, School end date, Enrol Status, Ground for Removal, Reason for leaving, Destination school, Exclusion reason, Exclusion start date, Exclusion end date, SEN Stage, SEN Needs, SEN History, Mode of travel, FSM History, Attendance, Student Service Family, Carer details, Carer address details, Carer contract details, Hearing Impairment And Visual Impairment, Education Psychology support, and Looked After status. (For in school children, the list is even longer, it lasts a lifetime, and it’s given away too).

Yet the Schools Bill would have granted Local Authorities powers to expand this already incredibly intrusive list to any further data at all of their choosing, without any limitation.

The CSJ perhaps most accurately, states in one report, that it is vulnerable children who are affected most by missing school time but this must not be conflated with Children Missing Education.

“In Autumn 2022, the latest term for which data is available, children in receipt of Free School Meals (FSM) had a severe absence rate which was triple the rate for children who were not eligible for FSM. Children in receipt of special educational needs (SEN) support are also more likely to be severely absent than their peers.”

Absenteeism and Children Missing Education are NOT the same. From the numbers above, I hope it is clear why.

The perception of reality matters in this topic area specifically because it is portrayed by the CSJ as an outcome of the pandemic. The CSJ is not politically neutral given its political founders, steering group and senior leadership with strong ties to the lockdown-skeptic COVID Recovery Group. That matters because it influences, and enables other influencers, to set the agenda on what is seen as cause and solution to a set of problems and the public policy interventions that are taken or funded as a result. In 2022 at the Tory party conference event on this subject which I also wrote up afterwards here, Iain Duncan Smith failed to acknowledge, even once, that thousands of people in the UK have died and continue to die or have lasting effects as a result of and with COVID-19.

It was such a contrast and welcome difference from the tone of Bridget Phillipson MPs speech today at CSJ, that she acknowledged what the pandemic reality was for thousands of families.

And after all, according to a King’s Fund report,Overall, the number of people who have died from Covid-19 to end-July 2022 is 180,000, about 1 in 8 of all deaths in England and Wales during the pandemic.” Furthermore in England and Wales, “The pandemic has resulted in about 139,000 excess deaths“. “Among comparator high-income countries (other than the US), only Spain and Italy had higher rates of excess mortality in the pandemic to mid-2021 than the UK.”

At the 2022 Conservative Conference fringe event, chaired by IDS, while there were several references made by the panel of the impact of the pandemic on children’s poor mental health, no one mentioned the cuts to youth services’ funding by 70% over ten years, that has allowed CAMHS funding and service provision to wither and fail children well before 2020. The pandemic exacerbated children’s pre-existing needs that the government has not only failed to meet since, but actively rationed and  reduced provision for. Event chair, Ian Duncan Smith, is also the architect of Universal Credit. And this matters in this very closely connected policy area for measuring and understanding the effectiveness of all these interventions.

Poverty and school attendance can but do not always have causes and correlations. But while we focus on the (inaccurately presented) number of children not in school we fail to pay attention to the Big Picture and conflated causes of children not in school. Missing bums on seats is not the problem, but a symptom. In some cases, literally. Historically, the main driver for absence is illness. In 2020/21, this was 2.1% across the full year. This was a reduction on the rates seen before the pandemic (2.5% in 2018/19).

What does persistent absentee mean?

Another part of these numbers often presented in the media is the ‘persistent absence’ rate. But is it meaningful? In 2018/19 the rate of persistent absentees (missing 10% of possible sessions, or the equivalent of one morning or one afternoon every week) was 10.8%. Now it is reported as around double that. But bear in mind that at any point in time, the label ‘persistent absentee’ may be misleading to the average man-on-the-street.

Don’t forget this is also a movable numbers game — understanding the definition of persistent absence and that it has changed three times since 2010, are both critical to appreciate the numbers and data used in discussing this subject.  Children are classified as persistently absent more quickly now than previously. The same number of children could be absent for the same number of school days as was a decade ago, but will be reported as having doubled.

A child who misses a full day in the term for unauthorised absence (illness) now stays marked as ‘persistently’ absent until their percentage of possible sessions in school outweighs the 10% missed.  So if they have 3.5 days of tummy bug or flu which every parent knows is pretty much the norm on “Back-to-School” then even the most dedicated pupils look ‘persistently’ absent for quite a while. It’s not only guilt inducing to the students who care, it creates stress to go-in-at-all-costs (including before recovery with risk of infection) for the ill, and stigmatises the disabled and those with long term health conditions.

A label unfit for purpose, it could usefully be re-named and the topic re-framed to rebuild trust with learners and families.

A table outlining the definitions by daily missed sessions to the persistent absence definition
Credit to George Stephenson High School, Newcastle upon Tyne. https://www.gshs.org.uk/attendance/what-persistent-absence-student

Another technical thing that could usefully be updated in terms of data collection is the inconsistency across Local Authorities for what age group they record which data. For example, children missing education (“CME”) is often Reception through to age 16 but for other categories it is for children to age 18, and age 25 for children with a special educational needs and disability (SEND) plan into young adulthood. Many Local Authorities use the definition in section 8 of the Education Act 1996 that is out of step with the more recently revised school leaving age in England.

Data-led decisions are not smart solutions

This is fundamentally not about data, but children’s lives.  In these debates there is a grave risk that a focus on the numbers, because of the way the data is presented, perceived, or used, means that reducing the numbers themselves becomes the goal.  The data is there. Joining it all up may feel like doing ‘something’ but it’s not going to contribute anything to getting bums on seats or deliver a quality education to every child. It won’t contribute to a solution except perhaps to some AI company CEO’s bottom line. And at what cost to children both by what you choose not to instead or direct harm? AI is not the solution or even a reliable tool when it comes to children’s social issues.

Entering data on a system with the hope of ‘spotting patterns’ without precise asks of data is rather like gazing at a crystal ball.  The computer cannot ‘guess’ what you are looking for. Rather than designing for the effective outcomes of what you want it to achieve, is symptomatic of the analysis of the problem. Lack of human authority and accountability.

It was said in the Laming report of the Victoria Climbié enquiry, there could be referrals coming in by fax, streaming on the floor and nobody picking them up. “It was not my job to pick up the fax from the fax machine. It was not my role. I had other things to do.” (5.22) All staff working in children and families’ services don’t need to be on one database to record data, but can work on decentralised systems with single role-based access. Systems can draw together data and present it without the need for a single record. Those are design questions, not a justification for building more databases and more national identifiers which may do nothing but duplicate the existing dysfunctions.

The Laming Review found that by the late 1990s social services had lost all of its human resources and training staff. They were overworked and missed things the computer showed them. Computers cannot make people do their job. Access to information does not create accountabilty for action.

In Victoria’s case, one example given was of a computer printout displaying a unique child reference number, that noted physical bruising.  The links were there for anyone who had access to see:

“but it seems likely that as the administrative staff were struggling to cope with the backlog of work at the time, it was simply overlooked.”

Children can experience the concentrated harms of poverty more than many in society. Some by government policy design. The data on child poverty is sometimes contradictory. But it is obvious to see that child poverty is not only more widespread but deeper now, than it was when the Conservative party took power in 2010. And it is important to consider a third further factor in the overlap between children’s school attendance and child poverty. Amos Toh, senior tech and human rights researcher at Human Rights Watch wrote recently on AI and public policy,

“As part of the welfare system since 2010, the government has ceded control of the country’s social assistance system to algorithms that often shrink people’s benefits in unpredictable ways. The system, known as Universal Credit, was rolled out in 2013 to improve administrative efficiency and save costs, primarily through automating benefits calculations and maximizing digital “self-service” by benefit claimants.”

Universal Credit has had a range of contested outcomes but  what should be uncontested is that the AI, the algorithms it uses, are flawed in various ways in various parts of the system.

So when I heard Shadow Education Secretary Bridget Phillipson say today that, “artificial intelligence (AI) will be used by Labour to spot absence trends to improve coordination between education, social care and the wider services that support families,” and, “plans to legislate for a new register of children in home education,” what I think to myself is this. Regardless of which political party is in power, imagine if there were a (duplicated) new database of children who are home educated and/ or known to be missing education at national level. Once these statistics are available at national level in one database (after all, only statistics not named records might be seen as necessary and proportionate beyond direct care, and practically the data will always be out of synch with local data), and imagine the money has not been spent on youth services, or Early Years intervention, but on ‘fix-it-all’ AI. What will change?

Fixing Britain is a People Problem

If you have followed the BBC Radio 4 Louise Casey ‘Fixing Britain’ series, you may or may not agree with all the suggestions but in the episode on Universal Credit, the summary is relevant for all of them. Often a public policy focus on money and legislation, forgets what it is about, people. “Policy disconnected from its purpose [people] is going to fail.” Policy makers often fail to understand most people’s lives whom the policy is intended to affect. Some of that today looks like this:

On poverty: There are far more food banks in the UK than branches of McDonalds.

On school leavers’ aspirations and opportunity: One third of children fail to get a pass in maths and English GCSE that is the gatekeeper to many jobs. AI-supported recruiting tools simply sort out and remove those who don’t have the qualifications in the applications process. Today one third of children are excluded from education and job opportunities not because they are necessarily unsuitable applicants, but because the grade boundaries are set so that one-third get D or lower.

On bad parenting: too often conflated into this debate by the Children’s Commissioner, children not-in-school is not a sign of bad parenting, any more than a child sent into school is a good one. Even joined up professional services and home visits can still be fobbed off and still fail to act on signs of neglect and abuse.  Parents and children on the radar of social services and on school rolls and in-school are known to the system and yet still it fails due to ‘underfunding of social services and the court system’.

On cuts to human support: Social workers vacancies were reported at a record high of 7,900 in 2022, a 21% rise on 2021. “The risks have been shown in safeguarding reviews after a series of scandals. A review of Bradford’s children’s services following Star Hobson’s death found record levels of vacancies and sickness among social workers.” No amount of data or Artificial intelligence can plug that hole in human capacity.

On policy aims: Much of this has been debated again and again. From twenty years ago, to the 2023 House of Commons Committee report on Persistent absence and support for disadvantaged pupils.

On children: Above all, contrary to some narratives, what is in a child’s best interests is not always being in school. If you ask primary children what they like and don’t like about school it may have changed little over time because what matters to them most is how it makes them feel. Some love sport, drama, music and art and are frustrated there is so little of it, and none at all from age 13 where the curriculum narrows to KS4 too early. For many it is not safe or supportive of their needs. Some are not fine in school and some need specialist support.  Expert individuals and oganisations identify those needs and are there to help. Children may more rarely be offered a choice or asked if they want to be in school, but without a consensual part in it, it doesn’t work.

On fault: Blame is then too often laid at parents’ feet, whether it is the narrative of parents are either feckless or failing to teach children to brush their teeth. In the year of a General Election how will this land with people who voted for the narrative, “Take back control”?

Control and choice

It was refreshing to hear Phillipson move at least a bit away from blame to responsibility and trust. The role of responsibility and trust in the system are, however, unevenly distributed and possibly under appreciated. Perhaps coincidentally, many parents and teachers today, are the first who went through the biggest costs and still have the largest debts owing from their Higher Education. As reported by politics.co.uk, “the results of the annual Higher Education Policy Unit and the Higher Education Academy student experience study in 2017 showed that just 35% of respondents believed their higher education experience represented ‘good’ or ‘very good’ value for money.” If parents see and act as if education is less of a gift in life or a public good, but more of a package that comes with consumer rights attached, then can you blame them? It was Labour that introduced the first student fees for Higher Ed.

It was the Conservatives who made ‘choice’ in the schools market central to their messaging on the role of parents in education for a decade. In the US they are now seeing the results of that ‘choice’ message, made even more extreme through per pupil cash transfers,  and by the political culture war divisions driven between communities and state schools that has helped steer state money away from the mainstream state school system.

Phillipson is right on why the current government approach isn’t working, “that broader reality is why the government’s approach – an Attendance Action Alliance – falls so far short of the challenge. Insofar as it tackles anything, it tackles the symptom, not the causes.”

But tackling things via different but wrong tools, won’t be better.

Failing to heed lessons from infrastructure projects, on pupil data, and AIs past and present, dooms us to repeat the same mistakes.  Who is in school is an outcome of the experience of the system at individual level, and if it delivers in the context of each child’s, family, and community life and the aims and quality of education. Focusing only on getting children into the classroom is of little value without understanding what the experience is like for them once there.  The outcomes of children not-in-school is not only a societal question, but one of long term sustainability for England’s state school system as a whole.


 

Waste products: bodily data and the datafied child

Recent conversations and the passage of the Data Protection and Digital Information Bill in parliament, have made me think once again about what the future vision for UK children’s data could be.

Some argue that processing and governance should be akin to a health model, first do no harm, professional standards, training, ISO lifecycle oversight, audits and governance bodies to approve exceptional releases and re-use.

Education data is health and body data

Children’s personal data in the educational context is remarkably often health data directly (social care, injury, accident, self harm, mental health) or indirectly (mood and emotion or eating patterns).

Children’s data in education is increasingly bodily data. An AI education company CEO was even reported to have considered, “bone-mapping software to track pupils’ emotions” linking a child’s bodily data and data of the mind. For a report written by Pippa King and myself in 2021, The State of Biometrics 2022: A Review of Policy and Practice in UK Education, we mapped the emerging prevalence of biometrics in educational settings. Published on the ten-year anniversary of the Protection of Freedoms Act 2012, we challenged the presumption that the data protection law is complied with well, or is effective enough alone in the protection of children’s data or digital rights.

We mustn’t forget, when talking about data in education, children do not go to school in order to produce data or to have their lives recorded, monitored or profiled through analytics. It’s not the purpose of their activity. They go to school to exercise their right in law to receive education, that data production is a by-product of the activity they are doing.

Education data as a by product of the process

Thinking of these together as children’s lives in by-products used by others, reminded me of the Alder Hey scandal published over twenty years ago, but going back decades.  In particular, the inquiry considered the huge store of body parts and residual human tissue of dead children accumulated between 1988 to 1995.

“It studied the obligation to establish ‘lack of objection’ in the event of a request to retain organs and tissue taken at a Coroner’s post-mortem for medical education and research.” (2001)

Thinking about the parallels of children’s personal data produced and extracted in education as a by-product, and organ and tissue waste a by-product of routine medical procedures in the living, highlights several lessons that we could be drawing today about digital processing of children’s lives in data and child/parental rights.

Digital bodies of the dead less protected than their physical parts

It also exposes gaps between the actual scenario today that the bodily tissue and the bodily data about deceased children could be being treated differently, since the data protection regime only applies to the living. We should really be forward looking and include rights here for all that go beyond the living “natural persons”, because our data does, and that may affect those who we leave behind. It is insufficient for researchers and others who wish to use data without restriction to object, because this merely pushes off the problem, increasing the risk of public rejection of ‘hidden’ plans later. (see DDM second reading briefing on recital 27, p 30/32),

What could we learn from handling body parts for the digital body?

In the children’s organ and tissue scandal, management failed to inform or provide suitable advice and support necessary to families.

Recommendations were made for change on consent to post-mortem examinations of children, and a new approach to consent and an NHS hospital post-mortem consent form for children and all residual tissue were adopted sector-wide.

The retention and the destruction of genetic material is considered in the parental consent process required for any testing that continues to use the bodily material from the child. In the Alder Hey debate this was about deceased children, but similar processes are in place now for obtaining parental consent to research re-use and retention for waste or ‘surplus’ tissue that comes from everyday operations on the living.

But new law in the Data Protection and Digital Information Bill is going to undermine current protections for genetic material in the future and has experts in that subject field extremely worried.

The DPDI Bill will consider the data of the dead for the first time

To date it only covers the data of or related to the living or “natural persons” and it is ironic that the rest of the Bill does the polar opposite, not about living and dead, but by redefining both personal data and research purposes it takes what is today personal data ‘in scope’ of data protection law and places it out of scope and beyond its governance due to exemptions, or changes in controller responsibility over time. Meaning a whole lot of data about children and the rest of us) will not be covered by DP law at all. (Yes, those are bad things in the Bill).

Separately, the new law as drafted, will also create a divergence from its generally accepted scope, and will start to bring data into scope the ‘personal data’ of the dead.

Perhaps as a result of limited parliamentary time, the DPDI Bill (see col. 939) is being used to include amendments on, “Retention of information by providers of internet services in connection with death of child,” to amend the Online Safety Act 2023 to enable OFCOM to give internet service providers a notice requiring them to retain information in connection with an investigation by a coroner (or, in Scotland, procurator fiscal) into the death of a child suspected to have taken their own life. The new clause also creates related offences.”

While primarily for the purposes of formal investigation into the role of social media in children’s suicide, and directions from Ofcom to social media companies to retain information for the period of one year beginning with the date of the notice, it highlights the difficulty of dealing with data after the death of a loved one.

This problem is perhaps no less acute where a child or adult has left no ‘digital handover’ via a legacy contact eg at Apple you can assign someone to be this person in the event of your own death from any cause. But what happens if your relation has not set this up and has been the holder of the digital key to your entire family photo history stored on a company’s cloud?  Is this a question of data protection, or digital identity management, or of physical product ownership?

Harvesting children’s digital bodies is not what people want

In our DDM research and report, “the words we use in data policy: putting people back in the picture” we explored how the language used to talk about personal data, has a profound effect on how people think about it.

In the current digital landscape personal data can often be seen as a commodity, a product to mine, extract and exploit and pass around to others. More of an ownership and IP question and the broadly U.S. approach. Data collection is excessive in “Big Data” mountains and “data lakes”, described just like the EU food surpluses of the 1970s. Extraction and use without effective controls creates toxic waste, is polluting and met with resistance. This environment is not sustainable and not what young people want. Enforcement of the data protection principles of purpose limitation and data minimisation should be helping here, but young people don’t see it.

When personal data is considered as ‘of the body’ or bodily residue, data as part of our life, the resulting view was that data is something that needs protecting. That need is generally held to be true, and represented in European human rights-based data laws and regulation. A key aim of protecting data is to protect the person.

In a workshop for that report preparation, teenagers expressed unease that data about them being ‘harvested’ to exploit as human capital and find their rights are not adequately enabled or respected. They find data can be used to replace conversation with them, and mean they are misrepresented by it, and at the same time there is a paradox that a piece of data can be your ‘life story’ and single source of truth advocating on your behalf.

Parental and children’s rights are grafted together and need recognised processes that respect this, as managed in health

Children’s competency and parental rights are grafted together in many areas of a child’s life and death, so why not by default in the digital environment? What additional mechanisms in a process are needed where both views carry legal weight? What are the specific challenges that need extra attention in data protection law due to the characteristics of data that can be about more than one person, be controlled by and not only be about the child, and parental rights?

What might we learn for the regulation of practice of a child’s digital footprint from how health manages residual tissue processing? Who is involved, what are the steps of the process and how is it communicated onwardly accompanying data flows around a system?

Where data protection rules do not apply, certain activities may still constitute an interference with Article 8 of the European Convention on Human Rights, which protects the right to private and family life. (WP 29 Opinion 4/2007 on the concept of personal data p24).

Undoubtedly the datafied child is an inseparable ‘data double’ of the child. Data users about children, who do so without their permission, without informing them or their families, without giving children and parents the tools to exercise their rights to have a say and control their digital footprint in life and in death, might soon find themselves being treated in the same way as accountable individuals in the Alder Hey scandal were, many years after the events took place.

 


Minor edits and section sub-headings added on 18/12 for clarity plus a reference to the WP29 opinion 04/2007 on personal data.

Automated suspicion is always on

In the Patrick Ness trilogy, Chaos Walking, the men can hear each others’ every thought but not the women.

That exposure of their bodily data and thought, means almost impossible privacy,  and no autonomy over their own bodily control of movement or of action. Any man that tries to block access to their thoughts is treated with automatic suspicion.

It has been on my mind since last week’s get together at FIPR. We were tasked before the event to present what we thought would be the greatest risk to rights [each pertinent to the speaker’s focus area] in the next five years.

Wendy Grossman said at the event and in her blog, “I’d look at the technologies being deployed around European and US borders to surveil migrants. Migrants make easy targets for this type of experimentation because they can’t afford to protest and can’t vote. “Automated suspicion,” Euronews.next calls it. That habit of mind is dangerous.” Those tools often focus on control of humans’ bodies. They infringe on freedom of movement.

In education, technology companies sell automated suspicion detection tools to combat plagiarism and cheating in exams. Mood detection to spot outliers in concentration. Facial detection to bar the excluded from premises or the lunch queue, or normalise behavioural anomalies, control physical attendance and mental presence. Automated suspicion is the opposite of building trusted human relationships.

I hadn’t had much space to think in the weeks before the event, between legislation, strategic litigation and overdue commitments to reports, events, and to others. But on reflection, I failed to explain why the topic area I picked above all others matters. It really matters.

It is the combination of a growth of children’s bodily data processing and SafetyTech deployed in schools. It’s not only because such tools normalise the surveillance of everything children do, send, share or search for on a screen, or that many enable the taking of covert webcam photos,  or even the profiles and labels it can create on terrorism and extremism or that can out LGBTQ+ teens. But that at its core, lies automated suspicion and automated control. Not only of bodily movement and actions, but of thought. Without any research or challenge to what that does to child development or their experience of social interactions and of authority.

First let’s take suspicion.

Suspicion of harms to self, harms to others, harms from others.

The software / systems / tools inspect the text or screen content the users enter into  devices (including text the users delete and text before it is encrypted) assuming a set of risks all of the time. When a potential risk is detected, the tools can capture and store a screenshot of the users’ screen. Depending on the company design and option bought, human company moderators may or may not first review the screenshots (recorded on a rolling basis also ‘without’ any trigger so as to have context ahead of the event) and text captures to verify the triggered events before sending to the school’s designated safeguarding lead. An estimated 1% of all triggered material might be sent on to a school to review and choose whether or not to act on. But regardless of that, the children’s data (including screenshots, text, and redacted text) may be stored for more than a year by the company before being deleted. Even content not seen as necessary but, “content which poses no risk on its own but is logged in case it becomes relevant in the future”.

Predictive threat, automated suspicion

In-school technology is not only capturing what is done by children but what they say they do, or might do, or think of doing. SafetyTech enables companies and school staff to police what children do and what they think, and it is quite plainly designed to intervene in actions and thoughts before things happen. It is predictively policing pupils in schools.

Safeguarding-in-schools systems were already one of my greatest emerging concerns but I suspect coinciding with recent wars, that the keywords in topics seen as connected to the Prevent programme will find a match rate at an all time high since 2016 and the risks it brings due to being wrong will have increased with it. But while we have now got various company CEOs talking about shared concerns, not least outing LGBTQ students as the CDT reported this year in the U.S. and a whistleblower who wanted to talk about the sensitive content the staff can see from their company side, there is not yet appetite to fix this across the sector. The ICO returned our case for sectoral attention, with no enforcement. DfE guidance still ignores the at home, out of hours contexts and those among the systems that can enable school staff or company staff to take photos of the children and no one might know. We’ve had lawyers write letters and submitted advice in consultations and yet it’s ignored to date.

Remember the fake bomb detectors that were golf ball machines? That’s the potential scenario we’ve got in education in “safeguarding in schools” tech. Automated decision making in black boxes that no one has publicly tested, no one can see inside, and we’ve no data on its discriminatory effects through language matching or any effective negative or false positives, and the harms it is or is not causing. We’ve risk averse institutions made vulnerable to scams. It may be utterly brilliant technology, with companies falling over independent testing that proves it ‘works’. I’ve just not seen any.

Some companies themselves say they need better guidance and agree there are significant gaps. Opendium, one leading provider of internet filtering and monitoring solutions, blogged about views expressed at a 2019 conference held by the Police Service’s Counter Terrorism Internet Referral Unit that schools need better advice .

Freedom of Thought

But it’s not just about what children do, but any mention of what they *might* do or their opinions of themselves, others or anything else. We have installed systems of thought surveillance into schools, looking for outliers or ‘extremists’ in different senses, and in its now everyday sense, underpinned by the Prevent programme and British Values. These systems do not only expose and create controls of children’s behaviours in what they do, but in their thoughts, their searches, what they type and share, send, or even, don’t and delete.

Susie Algere, human rights lawyer, describes, Freedom of Thought as, “protected absolutely in international human rights law. This means that, if an activity interferes with our right to think for ourselves inside our heads (the so-called “forum internum”) it can never be justified for any reason. The right includes three elements:

the right to keep our thoughts private
the right to keep our thoughts free from manipulation, and
the right not to be penalised for our thoughts.”

These SafetyTech systems don’t respect any of that. They infringe on freedom of thought.

Bodily data and contextual collapse

Depending on the company, SafetyTech may be built on keyword matching technology commonly used in the gaming tech industry.

Gaming data collected from children is a whole field in its own right – bodily data from haptics, and neuro data. Personal data from immersive environments that in another sector would be classified clearly as “health” data, and in the gaming sector too, will fall under the same “special category” or “sensitive data” due to its nature, not its context. But it is being collected at scale by companies that aren’t used to dealing with the demands of professional confidentiality and concept of ‘first do no harm’ that the health sector are founded on. Perhaps we’re not quite at the everyday for everyone in society, Ready Player One stage yet, but for those in communities who are creating a vast amount of data about themselves the questions over its oversight its retention, and perhaps its redistribution with authorities in particular with policing should be of urgent consideration. And those tools are on the way into the classroom.

At school level the enormous growth in the transfer of bodily data is not yet haptics but of bodily harm. A vast sector has grown up to support the digitisation of children’s safety, physical harms noticed by staff on children picked up at home, or accidents and incidents recorded at school. Often including marking full body outlines with where the injury has been.

The issues here again, are in part created by taking this data  beyond the physical environment of a child’s direct care and beyond the digital firewalls of child protection agencies and professionals. There are no clear universal policies on sealed records. ie not releasing the data of children-at-risk or those who undergo a name change, once it’s been added into school information management systems or into commercial company products like CPOMS, MyConcern, or Tootoot.

Similarly there is no clear national policy on the onward distribution into the National Pupil Database of the records of children in need (CiN) of child protection, which in my opinion, are inadequately shielded. The CIN census is a statutory social care data return made by every Local Authority to the Department for Education (DfE). It captures information about all children who have been referred to children’s social care regardless of whether further action is taken or not.

As of September 2022, there were only 70 individuals flagged for shielding and that includes both current and former pupils in the entire database. There were 23 shielded pupil records collected by the Department via the 2022 January censuses alone (covering early years, schools and alternative provision).

No statement or guidance is given direct to settings about excluding children from returns to the DfE. As of September 2022, there were 2,538,656 distinct CiN (any ‘child in need’ referred to children’s social care services within the year) / LAC ([state] looked after child) child records (going back to 2006), regardless of at-risk status, able to be matched to some home address information via other sources, (non CiN / LAC) all included in the NPD. The data is highly highly sensitive and detailed, including “categories of abuse” not only monitoring and capturing what has been done to children, but what is done by children.

Always on, always watching

The challenge for rights work in this sector is not primarily a technical problem but one of mindset. Do you think this is what schools are for? Are they aligned with the aims of education? One SafetyTech company CEO at a conference certainly marketed their tool as something that employers want children to get used to, to normalise the gaze of authority and monitoring of your attention span. In real Black Mirror stuff, you could almost hear him say, “their eyeballs belong to me for fifteen million merits”.

Monitoring in-class attendance is moving not only towards checking are you physically in school,  but are you present in focus as well.

Education is moving towards an always-on mindset for many, whether it be data monitoring and collection with the stated aims of personalising learning or the claims by companies that have trialed mood and emotion tech on pupils in England. Facial scanning is sold as a way of seeing if the class mood is “on point” with learning. Are they ‘engaged’?  After Pippa King spotted a live-trial in the wild starting in UK schools, we at Defend Digital Me had a chat with one company CEO who agreed after discussion, and the ICO blogpost on ’emotion tech’ hype, to stop that product rollout and cut it altogether from their portfolio. Under the EU AI Act it would soon be banned too, to protect children from its harms (children in the UK included, were Britain still under EU laws but now post-Brexit, they’re not).

The Times Education Commission reported in 2021 that Priya Lakhani told one of the Education Commission’s oral evidence sessions that Century Tech, “decided against using bone-mapping software to track pupils’ emotions through the cameras on their computers. Teachers were unhappy about pupils putting their cameras on for safeguarding reasons but there were also moral problems with supplying such technology to autocratic regimes around the world.”

But would you even consider this in an educational context at all?

Apps that blame and shame behaviours using RAG scores exposed to peers on wall projected charts are certainly already here. How long before such ’emotion’ and ‘mood’ tech emerges in Britain seeking a market beyond the ban in the EU, joined up with that which can blame and shame for lapses in concentration?

Is this simply the world now, that children are supposed to normalize third-party bodily surveillance and behavioural nudge?

That same kind of thinking in ‘estimation’ ‘safety’ and ‘blame’ might well be seen soon in eye scanning drivers in “advanced driver distraction warning systems”. Drivers staying ‘on track’ may be one area we will be expected to get used to monitoring our eyeballs, but will it be used to differentiate and discriminate between drivers for insurance purposes, or redirect blame for accidents? What about monitoring workers at computer desks, with smoking breaks and distraction costing you in your wage packet?

Body and Mind belong ‘on track’ and must be overseen

This routine monitoring of your face is expanding at pace in policing but policing the everyday to restrict access is going to affect the average person potentially far more than the use of facial detection and recognition in every public space. Your face is your passport and the computer can say no. Age as the gatekeeper of identity to participation and public and private spaces is already very much here online and will be expanded online in the UK by the Online Safety Act (noting other countries have realised its flaws and foolishness). Age verification and age assurance if given any weight, will inevitably lead to the balkanisation of the Internet, to throttling of content through prioritisation of who is permitted to do or see what, and control ofy content moderation.

In UK night clubs age verification is being normalised through facial recognition. Soon the only permitted Digital ID in what are (for now) purposes limited to rental and employment checks, will be the accredited government ID if the Data Protection and Digital Information Bill passes as drafted. But scope creep will inevitably move from what is possible, to what is required, across every aspect of our lives where identity is made an obligation for proof of eligibility.

Why all this matters is that we see a direction  of travel over and over again. Once “the data” is collected and retained there is an overwhelming desire down the line to say, well now we’ve got it, how can we use it? Increasingly that means joining it all up. And then passing it around to others. And the DPDI Bill takes away the safeguards around that over time (See KC opinion para 20, p.6).

It is something data protection law and lack of enforcement are already failing to protect us from adequately, because excessive data retention should be impossible under the data minimisation principle and purpose limitation, but controllers argue linked data ‘is not new data’. What we should see instead in enforcement is against the excessive retention of data that creates ‘new knowledge’ that goes beyond our reasonable expectations we see the government and companies gaining ever greater power to intervene in the lives of the data subjects, the people. The draft new law does the opposite.

Who decides what ‘on track’ looks like?

School SafetyTech is therefore the current embodiment of my greatest areas of concern for children’s rights in educational settings right now. Because it is an overlapping tech that monitors both what you do when, and claims to be able to put the thinking behind it in context. Tools in schools are moving towards prediction and interventions and the combinations of bodily control, thought, mood and emotion. They are shifting from on the server to on device and go with you everywhere your phone goes. ‘Interventions’ bring a whole new horizon of the potential infringements of rights and outcomes and questions of who decides what can be used for what purposes in a classroom, in loco parentis.

Filtering and monitoring technology in school “safetyTech”, blocks content and profiles the user over time. This monitoring of bodily behaviours, monitoring actions and thoughts, leads to staff acting on automated suspicion. It can lead to imposing control of bodily movement and of thoughts and actions. It’s adopted at scale for millions of children and students across the UK. It’s without oversight or published universal safety standards.

This is not a single technology, it’s a market and a mindset.

Who decides what is ‘suitable’, ‘on track’, and where ‘intervention’ is required is built into design?  It is not a problem of technology causing harm, but social and political choices and values embodied in technology that can be used to cause harm. For example in identifying and enabling the persecution of Muslim students that are fasting during Ramadan, based on their dining records. In the UK we have all the same tools already in place.

Who does any technology serve? is a question we have not yet resolved in education in England. The best interests of the child, the teacher, the institution, the State or company that built it?  Interests and incentives may overlap or may be contradictory. But who decides, and who is given the knowledge of how that was decided? As tech is becoming increasingly designed to run without any human intervention the effects of the automated decisions, in turn, can be significant, and happen at speed and scale.

Patrick Ness coined the phrase,”The Noise is a man unfiltered, and without a filter, a man is just chaos walking”. Controlling chaos may be a desirable government aim, but at what cost to whose freedoms?

AI in the public sector today, is the RAAC of the future

Reinforced Autoclaved Aerated Concrete (RAAC) used in the school environment is giving our Education Minister a headache. Having been the first to address the problem most publicly, she’s coming under fire as responsible for failure; for Ministerial failure to act on it in thirteen years of a Conservative government since 2010, and the failure of the fabric of educational settings itself.

Decades after buildings’ infrastructure started using RAAC, there is now a parallel digital infrastructure in educational settings. It’s worth thinking about what’s caused the RAAC problem and how it was identified. Could we avoid the same things in the digital environment and in the design, procurement and use of edTech products, and in particular, Artificial Intelligence?

Where has it been used?

In the procurement of school infrastructure, RAAC has been integrated into some parts of the everyday school system, especially in large flat roofs built around the 1960s-80s. It is now hard to detect and remedy or remove without significant effort. There was short-term thinking, short-term spending, and no strategy for its full life cycle or end-of-life expectations. It’s going to be expensive, slow, and difficult to find it and fix.

Where is the risk and what was the risk assessment?

Both most well-known recent cases, the 2016 Edinburgh School masonry collapse and the 2018 roof incident, happened in the early morning when no pupils were present, but, according to the 2019 safety alert by SCOSS, “in either case, the consequences could have been more severe, possibly resulting in injuries or fatalities. There is therefore a risk, although its extent is uncertain.”

That risk has been known for a long time, as today’s education minister Gillian Keegan rightly explained in that interview before airing her frustration. Perhaps it was not seen as a pressing priority because it was not seen as a new problem. In fact locally it often isn’t seen much at all, as it is either hidden behind front-end facades or built into hard-to-see places, like roofs. But already, ‘in the 1990s structural deficiencies became apparent’. (Discussed in papers by the Building Research Establishment (BRE) In the 1990s and again in 2002).

What has changed, according to expert reports, is that those visible problems are no longer behaving as expected in advance,  giving time for mitigation in what had previously been one-off catastrophic incidents. What was only affecting a few, could now affect the many at scale, and without warning. The most recent failures show there is no longer a reliable margin to act, before parts of the mainstream state education infrastructure pose children a threat to life.

Where is the similarity in the digital environment?

AI is the RAAC of another Minister’s future—it’s often similarly sold today as cost-saving, quick and easy to put in place.  You might need fewer people to install it rather than the available alternatives.

AI is being widely introduced at speed into children’s private and family life in England through its procurement and application in the infrastructure of public services; in education and children’s services and policing and in welfare; and some companies claim to be able to identify mood or autism or to be able to profile and influence mental health. Children rarely have any choice or agency to control its often untested effects or outcomes on them, in non-consensual settings.

If you’re working in AI “safety” right now, consider this a parable.

  • There are plenty of people pointing out risk in the current adoption of AI into UK public sector infrastructure; in schools, in health, in welfare, and in prisons and the justice system;
  • There are plenty of cases where harm is very real, but first seen by those in power as affecting the marginalised and minority;
  • There are no consistent published standards or obligations on transparency or of accountability to which AI sellers must hold their products before procurement and affect on people;
  • And there are no easily accessible records of where what type of AI is being procured and built into which public infrastructure, making tracing and remedy even harder in case of product recall.

The objectives of any company, State, service users, the public and investors may not be aligned. Do investors have a duty to ensure that artificial intelligence is developed in an ethical and responsible way? Prioritising short term economic gain and convenience, ahead of human impact or the long term public interest, has resulted in parts of schools’ infrastructure collapsing. And some AI is already going the same way.

The Cardiff Data Justice Lab together with Carnegie Trust have published numerous examples of cancelled systems across public services. “Pressure on public finances means that governments are trying to do more with less. Increasingly, policymakers are turning to technology to cut costs. But what if this technology doesn’t work as it should?” they asked.

In places where similar technology has been in place longer, we already see the impact and harm to people. In 2022, the Chicago Sun Times published an article noting that, “Illinois wisely stopped using algorithms in child welfare cases, but at least 26 states and Washington, D.C., have considered using them, and at least 11 have deployed them. A recent investigation found they are often unreliable and perpetuate racial disparities.” And the author wrote, “Government agencies that oversee child welfare should be prohibited from using algorithms.”

Where are the parallels in the problem and its fixes?

It’s also worth considering how AI can be “removed” or stopped from working in a system. Often not through removal at all, but simply throttling, shutting off that functionality. The problematic parts of the infrastructure remains in situ, but can’t easily be taken out after being designed-in. Whole products may also be difficult to remove.

The 2022 Institution of Structural Engineers’ report summarises the challenge now how to fix the current RAAC problems. Think about what this would mean doing to fix a failure of digital infrastructure:

  • Positive remedial supports and Emergency propping, to mitigate against known deficiencies or unknown/unproven conditions
  • Passive, fail safe supports, to mitigate catastrophic failure of the panels if a panel was to fail
  • Removal of individual panels and replacement with an alternative solution
  • Entire roof replacement to remove the ongoing liabilities
  • Periodic monitoring of the panels for their remaining service life

RAAC has not become a risk to life. It already was from design. While still recognised as a ‘good construction material for many purposes’ it has been widely used in unsafe ways in the wrong places.

RAAC planks made fifty years ago did not have the same level of quality control as we would demand today and yet was procured and put in place for decades after it was known to be unsafe for some uses, and risk assessments saying so.

RAAC was given an exemption from the commonly used codes of practice of reinforced concrete design (RC).

RAAC is scattered among non-RAAC infrastructure, making finding and fixing it, or its removal, very much harder than if it had been recorded in a register, making it easily traceable.

RAAC developers and sellers may no longer exist or have gone out of business without any accountability.

Current AI discourse should be asking not only for retrospective accountability or even life-cycle accountability, but also what does accountable AI look like by design and how do you guarantee it?

  • How do we prevent risk of harm to people from poor quality of systems designed to support them, what will protect people from being affected by unsafe products in those settings in the first place?
  • Are the incentives correct in procurement to enable adequate Risk Assessment be carried out by those who choose to use it?
  • Rather than accepting risk and retroactively expecting remedial action across all manner of public services in future—ignoring a growing number of ticking time bombs—what should public policy makers be doing to avoid putting them in place?
  • How will we know where the unsafe products were built into, if they are permitted then later found to be a threat-to-life?
  • How is safety or accountability upheld for the lifecycle of the product if companies stop making it, or go out of business?
  • How does anyone working with systems applied to people, assess their ongoing use and ensure it promotes human flourishing?

In the digital environment we still have margin to act, to ensure the safety of everyday parts of institutional digital infrastructure in mainstream state education and prevent harm to children. Whether that’s from parts of a product’s code, or use in the wrong way, or entire products. AI is already used in the infrastructure of school’ curriculum planning, curriculum content, or steering children’s self-beliefs and behaviours, and the values of the adult society these pupils will become. Some products have been oversold as AI when they weren’t, overhyped, overused and under explained,  their design is hidden away and kept from sight or independent scrutiny– some with real risks and harms. Right now, some companies and policy makers are making familiar errors and ‘safety-washing’ AI harms, ignoring criticism and pushing it off as someone else’s future problem.

In education, they could learn lessons from RAAC.


Background references

BBC Newsnight Timeline: reports from as far back as 1961 about aerated concrete concerns. 01/09/2023

BBC Radio 4 The World At One: Was RAAC mis-sold? 04/09/2023

Pre-1980 RAAC roof planks are now past their expected service life. CROSS. (2020) Failure of RAAC planks in schools.

A 2019 safety alert by SCOSS, “Failure of Reinforced Autoclaved Aerated Concrete (RAAC) Planks” following the sudden collapse of a school flat roof in 2018.

The Local Government Association (LGA) and the Department for Education (DfE) then contacted all school building owners and warned of ‘risk of sudden structural failure.’

In February 2022, the Institution of Structural Engineers published a report, Reinforced Autoclaved Aerated Concrete (RAAC) Panels Investigation and Assessment with follow up in April 2023, including a proposed approach to the classification of these risk factors and how these may impact on the proposed remediation and management of RAAC. (p.11)

image credit: DALL·E 2 OpenAI generated using the prompt “a model of Artificial Intelligence made from concrete slabs”.

 

Ensuring people have a say in future data governance

Based on a talk prepared for an event in parliament, hosted by Connected By Data and chaired by Lord Tim Clement-Jones, focusing on the Data Protection and Digital Information Bill, on Monday 5th December 17:00-19:00. “Ensuring people have a say in future data governance”.

Some reflections on Data in Schools (a) general issues; (b) the direction of travel the Government going in and; (c) what should happen, in the Bill or more widely.

Following Professor Sonia Livingstone who focussed primarily on the issues connected with edTech, I focussed on the historical and political context of where we are today, on ‘having a say’ in education data and its processing in, across, and out of, the public sector.


What should be different with or without this Bill?

Since I ran out of time yesterday I’m going to put first what I didn’t get around to: the key conclusions that point to what is possible with or without new Data Protection law. We should be better at enabling the realisation of existing data rights in the education sector today. The state and extended services could build tools for schools to help them act as controllers and for children to realise rights like a PEGE (a personalized exam grade explainer to show exam candidates what data was used to calculate their grade and how), Data usage reports should be made available at least annually from schools to help families understand what data about their children has gone where; and methods that enable the child or family to correct errors or express a Right to Object should be mandatory in schools’ information management systems.  Supplier standards on accuracy and error notifications should be made explicit and statutory, and supplier service level agreements affected by repeated failures.

Where is the change needed to create the social license for today’s practice, even before we look to the future?

“Ensuring people have a say in future data governance”. There has been a lot of asking lots of people for a say in the last decade. When asked, the majority of people generally want the same thingsboth those who are willing and less willing to have personal data about them re-used that was collected for administrative purposes in the public sectorto be told what data is collected for and how it is used, opt-in to re-use, to be able to control distribution, and protections for redress and against misuse strengthened in legislation.

Read Doteveryone’s public attitudes work. Or the Ipsos MORI polls or work by Wellcome. (see below). Or even the care.data summaries.

The red lines in the “Dialogues on Data” report from workshops carried out across different devolved regions of the UK for the 2013 ADRN remain valid today (about the reuse of deidentified linked public admin datasets by qualified researchers in safe settings not even raw identifying data), in particular with relation to:

  • Creating large databases containing many variables/data from a large number of public sector sources
  • Allowing administrative data to be linked with business data
  • Linking of passively collected administrative data, in particular geo-location data

“All of the above were seen as having potential privacy implications or allowing the possibility of reidentification of individuals within datasets. The other ‘red-line’ for some participants was allowing researchers for private companies to access data, either to deliver a public service or in order to make profit. Trust in private companies’ motivations were low.”

Much of this reflects what children and young people say as well. RAENG (2010) carried out engagement work with children on health data Privacy and Prejudice: young people’s views on the development and use of Electronic Patient Records (911.18 KB). They are very clear about wanting to keep their medical details under their own control and away from the ‘wrong hands’ which includes potential employers, commercial companies and parents.

Our own engagement work with a youth group aged 14-25 at a small scale was published in 2020 in our work, The Words We Use in Data Policy: Putting People Back in the Picture, and reflected what the Office for the Regulation of National Statistics went to publish in their own 2022 report, Visibility, Vulnerability and Voice (as a framework to explore whether the current statistics are helping society to understand the experiences of children and young people in all aspects of their lives). Young people worry about misrepresentation, about the data being used in place of conversations about them to take decisions that affect their lives, and about the power imbalance it creates without practical routes for complaint or redress. We all agree children’s voice is left out of the debate on data about them.

Parents are left out too. Defenddigitalme commissioned a parental survey via Survation (2018) under 50% felt they had sufficient control of their child’s digital footprint, and 2/3rds had not heard of the National Pupil Database or its commercial reuse.

So why is it that the public voice, loud and clear, is ignored in public policy and ignored in the drafting of the Data Protection and Digital Information Bill?

When it comes to education, debate should start with children’s and family rights in education, and education policy, not about data produced as its by-product.

The Universal Declaration of Human Rights Article 26 grafts a parent’s right onto child’s right to education, to choose the type of that education and it defines the purposes of education.

Education shall be directed to the full development of the human personality and to the strengthening of respect for human rights and fundamental freedoms. It shall promote understanding, tolerance and friendship among all nations, racial or religious groups, and shall further the activities of the United Nations for the maintenance of peace. Becoming a set of data points for product development or research is not the reason children go to school and hand over their personal details in the admissions process at all.

The State of the current landscape
To realise change, we must accept the current state of play and current practice. This includes a backdrop of trying to manage data well in the perilous state of public infrastructure, shrinking legal services and legal aid for children, ever-shrinking educational services in and beyond mainstream education, staff shortages and retention issues, and the lack of ongoing training or suitable and sustainable IT infrastructure for staff and learners.

Current institutional guidance and national data policy in the field is poor and takes the perspective of the educational setting but not the person.

Three key issues are problems from top-down and across systems:

  • Data repurposing i.e. SATS Key Stage 2 tests which are supposed to be measures of school performance not individual attainment are re-used as risk indicators in Local Authority datasets used to identify families for intervention, which it’s not designed for.
  • Vast amount of data distribution and linkage with other data: policing, economic drivers (LEO) and Local Authority broad data linkage without consent for purposes that exceed the original data collection purpose parents are told and use it like Kent, or Camden, “for profiling the needs of the 38,000 families across the borough”  plus further automated decision-making.
  • Accuracy in education data is a big issue, in part because families never get to see the majority of data created about a child much of which is opinion, and not submitted by them: ie the Welsh government fulfilled a Subject Access Request to one parent concerned with their own child’s record, and ended up revealing that every child in 2010 had been wrongly recorded thanks to a  Capita SIMS coding error, as having been in-care at some point in the past. Procurement processes should build penalties for systemic mistakes and lessons learned like this, into service level agreements, but instead we seem to allow the same issues to repeat over and over again.

What the DfE Does today

Government needs to embrace the fact it can only get data right, if it does the right thing. That includes policy that upholds the law by design. This needs change in its own purposes and practice.

National Pupil Data is a bad example from the top down. The ICO 2019-20 audit of the Department for Education — it is not yet published in full but findings included failings such as no Record of Processing Activity (ROPA), Not able to demonstrate compliance, and no fair processing. All of which will be undermined further by the Bill.

The Department for Education has been giving away 15 million people’s personal confidential data since 2012 and never told them. They know this. They choose to ignore it. And on top of that, didn’t inform people who were in school since then, that Mr Gove changed the law. So now over 21 million people’s pupil records are being given away to companies and other third parties, for use in ways we do not expect, and it is misused too. In 2015, more secret data sharing began, with the Home Office. And another pilot in 2018 with the DWP.

Government wanted to and changed the law on education admin data in 2012 and got it wrong. Education data alone is a sin bin of bad habits and complete lack of public and professional engagement, before even starting to address data quality and accuracy and backwards looking policy built on bad historic data.

The Commercial department do not have appropriate controls in place to protect personal data being processed on behalf of the DfE by data processors.” (ICO audit of the DfE , 2020)

Gambling companies ended up misusing access to learner records for over two years exposed in 2020 by journalists at the Sunday Times.

The government wanted nationality data from the Department for Education to be collected for the purposes of another (the Home Office) and got it very wrong. People boycotted the collection until it was killed off and data later destroyed.

Government changed the law on Higher Education in 2017 and got it wrong.  Now  third parties pass around named equality monitoring records like religion, sexual orientation, and disability and it is stored forever on named national pupil records. The Department for Education (DfE) now holds sexual orientation data on almost 3.2 million, and religious belief data on 3.7 million people.

After the summary findings published by the ICO of their compulsory audit of the Department for Education,  the question now is what will the Department and government do to address the 139 recommendations for improvement, with over 60% classified as urgent or high priority. Is the government intentional about change? We don’t think so at defend digital me, so we are, and welcome any support of our legal challenge.

Before we write new national law we must recognise and consider UK inconsistency and differences across education

Existing frameworks law and statutory guidance and recommendations need understood in the round (eg devolved education, including the age of a child and their capacity to undertake a contract in Scotland (at 16), the geographical applications of the Protection of Freedoms Act 2012, also the Prevent Duty since 2015 and its wider effects as a result of profiling children in counter-terrorism that reach beyond poor data protection and impacts on privacy (see The UN Special Rapporteur 2014 report on children’s rights and freedom of expression) – a plethora of Council of Europe work is applicable here in education that applies to UK as a member state: guidelines on data protection, AI, human rights, rule of law and the role of education in the promotion of democratic citizenship and a protection against authoritarian regimes and extreme nationalism.

The Bill itself
The fundamental principles of the GDPR and Data Protection law are undermined further from an already weak starting point since the 2018 Bill adopted exemptions that were not introduced by other countries in immigration and law enforcement.

  • The very definitions of personal and biometric data need close scrutiny.
  • Accountability is weakened (DPO, DPIA and prior consultation for high risk no longer necessary, ROPA)
  • Purpose limitation is weakened (legitimate interests and additional conditions for LI)
  • Redress is missing (Children and routes for child justice)
  • Henry VIII powers on customer data and business data must go.
  • And of course it only covers the living. What about children’s data misuse that causes distress and harms to human dignity but that is not covered strictly by UK Data Protection law, such as the children whose identities were used for undercover police in the SpyCops scandal. Recital 27 under the GDPR permits a possible change here.

Where are the Lessons Learned reflected in the Bill?

This Bill should be able to look at recent ICO enforcement action or judicial reviews to learn where and what is working and not working in data protection law. Lessons learned should be plentiful on public communications and fair processing, on the definitions of research, on discrimination, accuracy and bad data policy decisions. But where are those lessons in the Bill learned from health data sharing, why the care.data programme ran into trouble and similar failures repeated in the most recent GP patient data grab, or Google DeepMind and the RoyalFree? In policing from the Met Police Gangs Matrix?  In Home Affairs from the judicial review launched to challenge the lawfulness of an algorithm used by the Home Office to process visa applications? Or in education from the summer of 2020 exams fiasco?

The major data challenges as a result of government policy are not about data at all, but bad policy decisions which invariably mean data is involved because of ubiquitous digital first policy, public administration, and the nature of digital record keeping. In education examples include:

  • Partisan political agendas: i.e. the narrative of absence numbers makes no attempt to disaggregate the “expected” absence rate from anything on top, and presenting the idea as fact, that 100,000 children have not returned to school, “as a result of all of this”, is badly misleading to the point of being a lie.
  • Policy that ignores the law. The biggest driver of profiling children in the state education sector, despite the law that profiling children should not be routine, is the Progress 8 measure: about which Leckie & late Harvey Goldstein (2017) concluded in their work on the evolution of school league tables in England 1992-2016: ‘Contextual value-added’, ‘expected progress’ and ‘progress 8’ that, “all these progress measures and school league tables more generally should be viewed with far more scepticism and interpreted far more cautiously than have often been to date.”

The Direction of Travel
Can any new consultation or debate on the changes promised in data protection reform, ensure people have a say in future data governance, the topic for today, and what if any difference would it make?

Children’s voice and framing of children in National Data Strategy is woeful, either projected as victims or potential criminals. That must change.

Data protection law has existed in much similar form to today since 1984. Yet we have scant attention paid to it in ways that meet public expectations, fulfil parental and children’s expectations, or respect the basic principles of the law today. We have enabled technologies to enter into classrooms without any grasp of scale or risks in England that even Scotland has not with their Local Authority oversight and controls over procurement standards. Emerging technologies: tools that claim to be able to identify emotion and mood and use brain scanning, the adoption of e-proctoring, and mental health prediction apps which are treated very differently from they would be in the NHS Digital environment with ethical oversight and quality standards to meet — these are all in classrooms interfering with real children’s lives and development now, not some far-off imagined future.

This goes beyond data protection into procurement, standards, safety, understanding pedagogy, behavioural influence, and policy design and digital strategy. It is furthermore, naive to think this legislation, if it happens at all, is going to be the piece of law that promotes children’s rights when the others in play from the current government do not: the revision of the Human Rights Act, the recent PCSC Bill clauses on data sharing, and the widespread use of exemptions and excuses around data for immigration enforcement.

Conclusion
If policymakers who want more data usage treat people as producers of a commodity, and continue to ignore the publics’ “say in future data governance” then we’ll keep seeing the boycotts and the opt-outs and create mistrust in government as well as data conveners and controllers widening the data trust deficit**. The culture must change in education and other departments.

Overall, we must reconcile the focus of the UK national data strategy, with a rights-based governance framework to move forward the conversation in ways that work for the economy and research, and with the human flourishing of our future generations at its heart. Education data plays a critical role in social, economic, democratic and even security policy today and should be recognised as needing urgent and critical attention.


References:

Local Authority algorithms

The Data Justice Lab has researched how public services are increasingly automated and government institutions at different levels are using data systems and AI. However, our latest report, Automating Public Services: Learning from Cancelled Systems, looks at another current development: The cancellation of automated decision-making systems (ADS) that did not fulfil their goals, led to serious harm, or met caused significant opposition through community mobilization, investigative reporting, or legal action. The report provides the first comprehensive overview of systems being cancelled across western democracies.

New Research Report: Learning from Cancelled Systems

The Children of Covid: Where are they now? #CPC22

At Conservative Party Conference (“CPC22”) yesterday, the CSJ Think Tank hosted an event called, The Children of Lockdown: Where are they now?

When the speakers were finished, and other questions had been asked, I had the opportunity to raise the following three points.

They matter to me because I am concerned that bad policy-making for children will come from the misleading narrative based on bad data. The data used in the discussion is bad data for a number of reasons, based on our research over the last 4 years at defenddigitalme, and previously as part of the Counting Children coalition with particular regard to the Schools Bill.

The first is a false fact that has been often bandied about over the last year in the media and in Parliamentary debate, and that the Rt Hon Sir Iain Duncan Smith MP repeated in opening the panel discussion, that 100,000 children have not returned to school, “as a result of all of this“.

Full Fact has sought to correct this misrepresentation by individuals and institutions in the public domain several times, including one year ago today, when a Sunday Times article, published on 3 October 2021, claimed new figures showed “that between 95,000 and 135,000 children did not return to school in the autumn term, credited to the Commission on Young Lives, a task force headed up by former Children’s Commissioner for England.” Anne Longfield had then told Full Fact, that on 16 September 2021, “the rate of absence was around 1.5 percentage points higher than would normally be expected in the autumn term pre-pandemic.

Full Fact wrote, “This analysis attempts to highlight an estimated level of ‘unexplained absence’, and comes with a number of caveats—for example it is just one day’s data, and it does not record or estimate persistent absence.”

There was no attempt made in the CPC22 discussion to disaggregate the “expected” absence rate from anything on top, and presenting the idea as fact, that 100,000 children have not returned to school, “as a result of all of this”, is misleading.

Suggesting this causation for 100,000 children is wrong for two reasons. The first, is not talking about the number of children within that number who were out of school before the pandemic and reasons for that. The CSJ’s own report published in 2021, said that, “In the autumn term of 2019, i.e pre-Covid 60,244 pupils were labeled as severely absent.”

Whether it is the same children or not who were out of school before and afterwards also matters to apply causation. This named pupil-level absence data is already available for every school child at national level on a termly basis, alongside the other personal details collected termly in the school census, among other collections.

Full Fact went on to say, “The Telegraph reported in April 2021 that more than 20,000 children had “fallen off” school registers when the Autumn 2020 term began. The Association of Directors of Children’s Services projected that, as of October 2020, more than 75,000 children were being educated at home. However, as explained above, this is not the same as being persistently absent.”

The second point I made yesterday, was that the definition of persistent absence has changed three times since 2010, so that children are classified as persistently absent more quickly now at 10%, than when it meant 20% or more of sessions were missed.

(It’s also worth noting that data are inconsistent over time in another way too. The 2019 Guide to Absence Statistics draws attention to the fact that, “Year on year comparisons of local authority data may be affected by schools converting to academies.”)

And third and finally, I pointed out where we have found a further problem in counting children correctly. Local Authorities do this in different ways. Some count each actual child once in the year in their data, some count each time a child changes status (i.e a move from mainstream into Alternative Provision to Elective Home Education could see the same child counted three times in total, once in each dataset across the same year), and some count full-time equivalent funded places (i.e. if five children each have one day a week outside mainstream education, they would be counted only as one single full-time child in total in the reported data).

Put together, this all means not only that the counts are wrong, but the very idea of “ghost children” who simply ‘disappear’ from school without anything known about them anywhere at all, is a fictitious and misleading presentation.

All schools (including academies and independent schools) must notify their local authority when they are about to remove a pupil’s name from the school admission register under any of the fifteen grounds listed in Regulation 8(1) a-n of the Education (Pupil Registration) (England) Regulations 2006. On top of that, children are recorded as Children Missing Education, “CME” where the Local Authority decides a child is not in receipt of suitable education.

For those children,  processing of personal data of children not-in-school by Local Authorities is already required under s436Aof the The Education Act 1996, Duty to make arrangements to identify children not receiving education.

Research done as part of the Counting Children coalition with regards to the Schools Bill, has found every Local Authority that has replied to date (with a 67% response rate to FOI on July 5, 2022) upholds its statutory duty to record these children who either leave state education, or who are found to be otherwise missing education. Every Local Authority has a record of these children, by name, together with much more detailed data.**  The GB News journalist on the panel said she had taken her children out of school and the Local Authority had not contacted her. But as a home-educating audience member then pointed out, that does not mean therefore the LA did not know about her decision, since they would already have her child-/ren’s details recorded. There is law in place already on what LAs must track. Whether or not and how the LA is doing its job, was beyond this discussion, but the suggestion that more law is needed to make them collect the same data as is already required is superfluous.

This is not only about the detail of context and nuance in the numbers and its debate, but substantially alters the understanding of the facts. This matters to have correct, so that bad policy doesn’t get made based on bad data and misunderstanding the conflated causes.

Despite this, in closing Iain Duncan Smith asked the attendees to go out from the meeting and evangelise about these issues. If they do so based on his selection of ‘facts’ they will spread misinformation.

At the event, I did not mention two further parts of this context that matter if policy makers and the public are to find solutions to what is no doubt an important series of problems, and that must not be manipulated to present as if they are entirely as a result of the pandemic. And not only the pandemic, but lockdowns specifically.

Historically, the main driver for absence is illness. In 2020/21, this was 2.1% across the full year. This was a reduction on the rates seen before the pandemic (2.5% in 2018/19).

A pupil on-roll is identified as a persistent absentee if they miss 10% or more of their possible sessions (one school day has two sessions, morning and afternoon.)  1.1% of pupil enrolments missed 50% or more of their possible sessions in 2020/21. Children with additional educational and health needs or disability, have higher rates of absence. During Covid, the absence rate for pupils with an EHC plan was 13.1% across 2020/21.

Authorised other reasons has risen to 0.9% from 0.3%, reflecting that vulnerable children were prioritised to continue attending school but where parents did not want their child to attend, schools were expected to authorise the absence.” (DfE data, academic year 2020/21)

While there were several references made by the panel to the impact of the pandemic on children’s poor mental health, no one mentioned the cuts to youth services’ funding by 70% over ten years, that has allowed CAMHS funding and service provision to wither and fail children well before 2020. The pandemic has exacerbated children’s pre-existing needs that the government has not only failed to meet since, but actively reduced provision for.

It was further frustrating to hear, as someone with Swedish relatives, of their pandemic approach presented as comparable with the UK and that in effect, they managed it ‘better’. It seems absurd to me, to compare the UK uncritically with a country with the population density of Sweden. But if we *are* going to do comparisons with other countries, it should be with fuller understanding of context, and all of their data, and caveats if comparison is to be meaningful.

I was somewhat surprised that Iain Duncan Smith also failed to acknowledge, even once, that thousands of people in the UK have died and continue to die or have lasting effects as a result of and with COVID-19. According to the King’s Fund report,Overall, the number of people who have died from Covid-19 to end-July 2022 is 180,000, about 1 in 8 of all deaths in England and Wales during the pandemic.” Furthermore in England and Wales, “The pandemic has resulted in about 139,000 excess deaths“. “Among comparator high-income countries (other than the US), only Spain and Italy had higher rates of excess mortality in the pandemic to mid-2021 than the UK.” I believe that if we’re going to compare ‘lockdown success’ at all, we should look at the wider comparable data before making it. He might also have chosen to mention alongside this, the UK success story of research and discovery, and the NHS vaccination programme.

And there was no mention at all made of the further context, that while much was made of the economic harm of the impact of the pandemic on children, “The Children of Lockdown” are also, “The Children of Brexit”. It is non-partisan to point out this fact, and, I would suggest, disingenuous to leave out entirely in any discussion of the reasons for or impact of economic downturn in the UK in the last three years. In fact, the FT recently called it a “deafening silence.”

At defenddigitalme, we raised the problem of this inaccurate “counting” narrative numerous times including with MPs, members of the House of Lords in the Schools Bill debate as part of the Counting Children coalition, and in a letter to The Telegraph in March this year. More detail is here, in a blog from April.


Update May 23, 2023

Today I received the DfE held figures of he number of children who leave an educational setting for an unknown onward destination, a section of the Common Transfer Files holding space, in effect a digital limbo after leaving an educational setting until the child is ‘claimed’ by the destination. It’s  known as, the Lost Pupils Database.

Furthermore, the DfE has published exploratory statistics on EHE
and ad hoc stats on CME too.

October 2022. More background:

The panel was chaired by the Rt Hon Sir Iain Duncan Smith MP and other speakers included Fraser Nelson, Editor of The Spectator Magazine; Kieron Boyle, Chief Executive Officer of Guy’s & St Thomas Foundation; the Rt Hon Robert Halfon MP, Education Select Committee Chair; and Mercy Muroki, Journalist at GB News.

We have previously offered to share our original research data and discuss with the Department for Education, and repeated this offer to the panel to help correct the false facts. I look forward in the hope they will take it up.

** Data collected in the record by Local Authorities when children are deregistered from state education (including to move to private school) may include a wide range of personal details, including as an example in Harrow: Family Name, Forename, Middle name, DOB, Unique Pupil Number (“UPN”), Former UPN, Unique Learner Number, Home Address (multi-field), Chosen surname, Chosen given name, NCY (year group), Gender, Ethnicity, Ethnicity source, Home Language, First Language, EAL (English as an additional language), Religion, Medical flag, Connexions Assent, School name, School start date, School end date, Enrol Status, Ground for Removal, Reason for leaving, Destination school, Exclusion reason, Exclusion start date, Exclusion end date, SEN Stage, SEN Needs, SEN History, Mode of travel, FSM History, Attendance, Student Service Family, Carer details, Carer address details, Carer contract details, Hearing Impairment And Visual Impairment, Education Psychology support, and Looked After status.

Policing thoughts, proactive technology, and the Online Safety Bill

Former counter-terrorism police chief attacks Rishi Sunak’s Prevent plans“, reads a headline in today’s Guardian. “Former counter-terrorism chief Sir Peter Fahy […] said: “The widening of Prevent could damage its credibility and reputation. It makes it more about people’s thoughts and opinions. Fahy said: “The danger is the perception it creates that teachers and health workers are involved in state surveillance.”

This article leaves out that today’s reality is already far ahead of proposals or perception. School children and staff are already surveilled in these ways. Not only are things monitored that people think type or read or search for online and offline in the digital environment, but copies may be collected, retained by companies and interventions made.

The products don’t only permit monitoring of trends on aggregated data in overviews of student activity but the behaviours of individual students. And these can be deeply intrusive and sensitive when you are talking about self harm, abuse, and terrorism.

(For more on the safety tech sector, often using AI in proactive monitoring, see my previous post (May 2021) The Rise of Safety Tech.)

Intrusion through inference and interventions

From 1 July 2015 all schools have been subject to the Prevent duty under section 26 of the Counter-Terrorism and Security Act 2015, in the exercise of their functions, to have “due regard to the need to prevent people from being drawn into terrorism”.  While these products are about monitoring far more than the remit of Prevent,  many companies actively market online filtering, blocking and monitoring safety products as a way of meeting that in the digital environment. Such as, “Lightspeed Filter™ helps you meet all of the Prevent Duty’s online regulations…

Despite there being no obligation to date, to fulfil this duty through technology, some companies’ way of selling such tools could be interpreted as threatening if schools don’t use it. Like this example:

“Failure to comply with the requirements may result in intervention from the Prevent Oversight Board, prompt an Ofsted inspection or incur loss of funding.”

Such products may create and send real-time alerts to company or school staff when children attempt to reach sites or type “flagged words” related to radicalisation or extremism on any online platform.

Under the auspices of the safeguarding-in-schools data sharing and web monitoring in the Prevent programme children may be labelled with terrorism or extremism labels, data which may be passed on to others or stored outside the UK without their knowledge. The drift in what is considered significant, has been from terrorism into now more vague and broad terms of extremism and radicalisation. Away from some assessment of intent and capability of action, into interception and interventions for potentially insignificant potential vulnerabilities and inferred assumptions of disposition towards such ideas. This is not potentially going to police thoughts as suggested by Fahy of Sunak’s views. It is already doing so. Policing thoughts in the developing child and holding them accountable for it like this in ways that are unforeseeable, is inappropriate and requires thorough investigation into its effects on children, including mental health.

But it’s important to understand that these libraries of thousands of words, ever changing and in multiple languages, and what the systems are looking for and flag, often claiming to do it using Artificial Intelligence, go far beyond Prevent. ‘Legal but harmful’ is their bread and butter. Self harm, harm to or from others.

While companies have no obligations to publish how the monitoring or flagging operates, what the words or phrases or blocked websites are, their error rates (positive and negative) or the effects on children or school staff and their behaviour as a result, these companies have a great deal of influence what gets inferred from what children do online, and who decides what to act on.

Why does it matter?

Schools have normalized the premise that systems they introduce should monitor activity outside of the school network, and hours. And that strangers or their private companies’ automated systems should be involved in inferring or deciding what children are ‘up to’ before the school staff who know the children in front of them.

In a defenddigitalme report, The State of Data 2020, we included a case study on one company that has since been bought out.  And bought again. As of August 2018 eSafe was monitoring approximately one million school children plus staff across the UK. This case study they used in their public marketing raised all sorts of questions on professional  confidentiality and school boundaries, personal privacy, ethics, and companies’ role and technical capability, as well as the lack of any safety tech accountability.

“A female student had been writing an emotionally charged letter to her Mum using Microsoft Word, in which she revealed she’d been raped. Despite the device used being offline, eSafe picked this up and alerted John and his care team who were able to quickly intervene.”

Their then CEO  had told the House of Lords 2016 Communication Committee enquiry on the Children and the Internet, how the products are not only monitoring children in school or school hours:

“Bearing in mind we are doing this throughout the year, the behaviours we detect are not confined to the school bell starting in the morning and ringing in the afternoon, clearly; it is 24/7 and it is every day of the year. Lots of our incidents are escalated through activity on evenings, weekends and school holidays.”

Similar products offer a photo capturing feature of users (pupils while using the device being monitored) described as “common across most solutions in the sector” by this company:

When a critical safeguarding keyword is copied, typed or searched for across the school network, schools can turn on NetSupport DNA’s webcams capture feature (this feature is turned-off by default) to capture an image of the user (not a recording) who has triggered the keyword.

How many webcam photos have been taken of children by school staff or others through those systems, and for what purposes, kept by whom? In the U.S. in 2010, Lower Merion School District, Philadelphia settled a lawsuit for using laptop webcams to take photos of students.  Thousands of photos had been taken even at home, out of hours, without their knowledge.

Who decides what does and does not trigger interventions across different products? In the month of December 2017 alone, eSafe claims they added 2254 words to their threat libraries.

Famously, Impero’s system even included the word “biscuit” which they say is a term used to define a gun. Their system was used by more than “half a million students and staff in the UK” in 2018. And students had better not talk about “taking a wonderful bath.” Currently there is no understanding or oversight of the accuracy of this kind of software and black-box decision-making is often trusted without openness to human question or correction.

Aside from how the range of tools that are all different work, there are very basic questions about whether such policies and tools help or harm children in various ways at all. The UN Special Rapporteur’s 2014 report on children’s rights and freedom of expression stated:

“The result of vague and broad definitions of harmful information, for example in determining how to set Internet filters, can prevent children from gaining access to information that can support them to make informed choices, including honest, objective and age-appropriate information about issues such as sex education and drug use. This may exacerbate rather than diminish children’s vulnerability to risk.” (2014)

U.S. safety tech creates harms

Today in the U.S. the CDT published a report on school monitoring systems there, many of which are also used over here. The report revealed that 13 percent of students knew someone who had been outed as a result of student-monitoring software. Another conclusion the CDT draws, is that monitoring is used for discipline more often than for student safety.

We don’t have that same research for the UK, but we’ve seen IT staff openly admit to using the webcam feature to take photos of young boys who are “mucking about” on the school library computer.

The Online Safety Bill scales up problems like this

The Online Safety Bill seeks to expand how such ‘behavioural identification technology’ can be expanded outside schools.

“Proactive technology include content moderation technology, user profiling technology or behaviour identification technology which utilises artificial intelligence or machine learning.” (p151 Online Safety Bill, August 3, 2022)

The “proactive technology requirement” is as yet rather open ended, left to OFCOM in Codes of Practice but the scope creep of such AI-based tools has become ever more intrusive in education. Legal but harmful is decided by companies and the IWF and any number of opaque third parties whose process and decision-making we know little about. It’s important not to conflate filtering, blocking lists of ‘unsuitable’ websites that can be accessed in schools, with monitoring and tracking individual behaviours.

‘Technological developments that have the capacity to interfere with our freedom of thought fall clearly within the scope of “morally unacceptable harm,”‘ according to Algere (2017), and yet this individual interference is at the very core of school safeguarding tech and policy by design.

In 2018, the ‘lawful but harmful’ list of activities in the Online Harms White paper was nearly identical with those terms used by school Safety Tech companies. The Bill now appears to be trying to create a new legitimate basis for these practices, more about underpinning a developing market, than supporting children’s safety or rights.

Chilling speech is itself controlling content

While a lot of debate about the Bill has been the free speech impacts of content removal, there has been less about what is unwritten but how it will operate to prevent speech and participation in the digital environment for children. The chilling effect of surveillance on access and participation online is well documented. Younger people and women are more likely to be negatively affected (Penney, 2017). The chilling effect on thought and opinion is worsened in these types of tools that trigger an alert even when what is typed is quickly deleted or remains unsent or shared. Thoughts are no longer private.

The ability to use end-to-end encryption on private messaging platforms is simply worked around by these kinds of tools, trading security for claims of children’s safety. Anything on screen may be read in the clear by some systems, even capturing passwords and bank details.

Graham Smith has written, “It may seem like overwrought hyperbole to suggest that the [Online Harms] Bill lays waste to several hundred years of fundamental procedural protections for speech. But consider that the presumption against prior restraint appeared in Blackstone’s Commentaries (1769). It endures today in human rights law. That presumption is overturned by legal duties that require proactive monitoring and removal before an independent tribunal has made any determination of illegality.”

More than this, there is no determination of illegality in legal but harmful activity. It’s opinion. The government is prone to argue that, “nothing in the Bill says X…” but you need to understand this context of how such proactive behavioural monitoring tools work is through threat and the resultant chilling effect to impose unwritten control. This Bill does not create a safer digital environment, it creates threat models for users and companies, to control how we think and behave.

What do children and parents think?

Young people’s own views that don’t fit the online harms narrative have been ignored by Westminster scrutiny Committees. A 2019 survey by the Australian e-safety commissioner found that over half (57%) of child respondents were uncomfortable with background monitoring processes, and 43 %were unsure about these tools’ effectiveness in ensuring online safety.

And what of the role of parents? Article 3(2) of the UNCRC says: “States Parties undertake to ensure the child such protection and care as is necessary for his or her wellbeing, taking into account the rights and duties of his or her parents, legal guardians, or other individuals  legally responsible for him or her, and, to this end, shall take all appropriate legislative and administrative measures.” (my emphasis)

In 2018, 84% of 1,004 parents in England who we polled through Survation, agreed that children and guardians should be informed how this monitoring activity works and wanted to know what the keywords were. (We didn’t ask if it should happen at all or not.)

The wide ranging nature [of general monitoring] rather than targeted and proportionate interference has been judged to be in breach of law and a serious interference with rights, previously. Neither policy makers nor companies should assume parents want safety tech companies to remove autonomy, or make inferences about our children’s lives. Parents if asked, reject the secrecy in which it happens today and demand transparency and accountability. Teachers can feel anxious talking about it at all. There’s no clear routes for error corrections, in fact it’s not done because some claim in building up profiles staff should not delete anything and ignore claims of errors, in case a pattern of behaviour is missed. But there’s no independent assessments available to evidence these tools work or are worth the costs. There are no routes for redress or responsibility taken for tech-made mistakes. None of which makes children safer online.

Before broadening out where such monitoring tools are used, their use and effects on school children need to be understood and openly debated. Policy makers may justify turning a blind eye to harms created by one set of technology providers while claiming that only the other tech providers are the problem, because it suits political agendas or industry aims, but children’s rights and their wellbeing should not be sacrificed in doing so.  Opaque, unlawful and unsafe practice must stop. A quid pro quo for getting access to millions of children’s intimate behaviour, should be transparent access to their product workings, and accepting standards on universal safe accountable practices. Families need to know what’s recorded. To have routes for redress when a daughter researching ‘cliff walks’ gets flagged as a suicide risk or an environmentally interested teenage son searching for information on ‘black rhinos’ is asked about his potential gang membership. The tools sold as solutions to online harms, shouldn’t create more harm like these reported real-life case studies.

Teachers are ‘involved in state surveillance’ as Fahy put it, through Prevent. Sunak was wrong to point away from the threats of the far right in his comments. But the far broader unspoken surveillance of children’s personal lives, behaviours and thoughts through general monitoring in schools, and what will be imposed through the Online Safety Bill more broadly, should concern us far more than was said.

On #IWD2022 gender bias in #edTech

I’m a mother of three girls at secondary school. For international women’s day 2022 I’ve been thinking about the role of school technology in my life.

Could some of it be improved to stop baking-in gender discrimination norms to home-school relationships?

Families come in all shapes and sizes and not every family has defined Mum and Dad roles. I wonder if edTech could be better at supporting families if it offered the choice of a multi-parent-per-child relationship by-default?

School-home communications rarely come home in school bags anymore, but digitally, and routinely sent to one-parent-per-child. If something needs actioned, it’s typically going to one parent, not both. The design of digital tools can lock-in the responsibility for action to a single nominated person. Schools send the edTech company the ‘pupil parent contact’ email, but, at least in my experience, don’t ever ask what that should be after it’s been collected once. (And don’t do a good job of communicating data rights each time before doing so either, but that’s another story.)

Whether it’s about learning updates with report cards about the child, or weekly newsletters, changes of school clubs, closures, events or other ‘things you should know’ I filter emails I get daily from a number of different email accounts for relevance, and forward them on to Dad.

To administer cashless payments to school for contributions to art, cooking, science and technology lessons, school trips, other extras or to manage my child’s lunch money, there is a single email log-in and password for a parent role allocated to the child’s account.

And it might be just my own unrepresentative circle of friends, but it’s usually Mum who’s on the receiving end of demands at all hours.

In case of illness, work commitments, otherwise being unable to carry on as usual, it’s no longer as easy for a second designated parent role to automatically pick up or share the responsibilities.

One common cashless payment system’s approach does permit more than one parent role, but it’s manual and awkward to set up. “For a second parent to have access it is necessary for the school to send a second letter with a second temporary username and password combo to activate a second account. In short, the only way to do this is to ask your school.”

Some messaging services allow a school-to-multiple-parent email, but the message itself often forms an individual not group thread with the teacher, i.e designed for a class not a family.

Some might suggest it is easy enough to set up automatic email forwarding, but again this pushes back the onus onto the parent and doesn’t solve the problem of only one person able to perform transactions.

I wonder if one-way communications tools offered a second email address by default what difference it would make to overall parental engagement?

What if for financial management edTech permitted an option to have a ‘temporary re-route’ to another email address, or default second role with notification to the other something had been paid?

Why can’t one parent, once confirmed with secure access to the child-parent account, add a second parent role? These need not be the parent, but another relation managing the outgoing money. You can only make outgoing payments to the school, or withdraw money to the same single bank account it comes from, so fraud isn’t likely.

I wonder what research would look like at each of these tools, to assess whether there is a gender divide built into default admin?

What could it improve in work-life balance for staff and families, if emails were restricted to send or receive in preferred time windows?

Technology can be amazing and genuinely make life easier for some. But not everyone fits the default and I believe the defaults are rarely built to best suit users, but rather the institutions that procure them. In many cases edTech aren’t working well for the parents that make up their main user base.

If I were designing these, they’d be school not third-party cloud based, and distributed systems, centred on the child. I think we can do better, not only for women, but everyone.


PS When my children come home from school today, I’ll be showing them the Gender Pay Gap Bot @PayGapApp thread with explanations of mode, mean and median and worth a look.

Man or machine: who shapes my child? #WorldChildrensDay 2021

A reflection for World Children’s Day 2021. In ten years’ time my three children will be in their twenties. What will they and the world around them have become? What will shape them in the years in between?


Today when people talk about AI, we hear fears of consciousness in AI. We see, I, Robot.  The reality of any AI that will touch their lives in the next ten years is very different. The definition may be contested but artificial intelligence in schools already involves automated decision making at speed and scale, without compassion or conscience, but with outcomes that affect children’s lives for a long time.

The guidance of today—in policy documents, and well intentioned toolkits and guidelines and oh yes yet another ‘ethics’ framework— is all fairly same-y in terms of the issues identified.

Bias in training data. Discrimination in outcomes. Inequitable access or treatment. Lack of understandability or transparency of decision-making. Lack of routes for redress. More rarely thoughts on exclusion, disability and accessible design, and the digital divide. In seeking to fill it, the call can conclude with a cry to ensure ‘AI for all’.

Most of these issues fail to address the key questions in my mind, with regards to AI in education.

Who gets to shape a child’s life and the environment they grow up in? The special case of children is often used for special pleading in government tech issues. Despite this, in policy discussion and documents, govt. fails over and over again to address children as human beings.

Children are still developing. Physically, emotionally, their sense of fairness and justice, of humor, of politics and who they are.

AI is shaping children in ways that schools and parents cannot see.  And the issues go beyond limited agency and autonomy. Beyond the UNCRC articles 8 and 18, the role of the parent and lost boundaries between schools and home, and 23 and 29. (See at the end in detail).

Concerns about accessibility published on AI are often about the individual and inclusion, in terms of design to be able to participate. But once they can participate, where is the independent measurement and evaluation of impact on their educational progress, or physical and mental development? What is their effect?

From overhyped like Edgenuity, to the oversold like ClassCharts (that didn’t actually have any AI in it but it still won Bett Show Awards), frameworks often mention but still have no meaningful solutions for the products that don’t work and fail.

But what about the harms from products that work as intended? These can fail human dignity or create a chilling effect, like exam proctoring tech. Those safety tech that infer things and cause staff to intervene even if the child was only chatting about ‘a terraced house.’ Punitive systems that keep profiles of behaviour points long after a teacher would have let it go. What about those shaping the developing child’s emotions and state of mind by design and claim to operate within data protection law? Those who measure and track mental health or make predictions for interventions by school staff?

Brain headbands to transfer neurosignals aren’t biometric data in data protection terms if not used to or able to uniquely identify a child.

“Wellbeing” apps are not being regulated as medical devices and yet are designed to profile and influence mental health and mood and schools adopt them at scale.

If AI is being used to deliver a child’s education, but only in the English language, what risk does this tech-colonialism create in evangelising  children in non-native English speaking families through AI, not only in access to teaching, but on reshaping culture and identity?

At the institutional level, concerns are only addressed after the fact. But how should they be assessed as part of procurement when many AI are marketed as , it never stops “learning about your child”? Tech needs full life-cycle oversight, but what companies claim their products do is often only assessed to pass accreditation at a single point in time.

But the biggest gap in governance is not going to be fixed by audits or accreditation of algorithmic fairness. It is the failure to recognize the redistribution of not only agency but authority; from individuals to companies (teacher doesn’t decide what you do next, the computer does). From public interest institutions to companies (company X determines the curriculum content, not the school). And from State to companies (accountability for outcomes has fallen through the gap in outsourcing activity to the AI company). We are automating authority, and with it the shirking of responsibility, the liability for the machine’s flaws, and accepting it is the only way, thanks to our automation bias. Accountability must be human, but whose?

Around the world the rush to regulate AI, or related tech in Online Harms, or Digital Services, or Biometrics law, is going to embed, not redistribute power, through regulatory capitalism.

We have regulatory capture including on government boards and bodies that shape the agenda; unrealistic expectations of competition shaping the market; and we’re ignoring transnational colonialisation of whole schools or even regions and countries shaping the delivery of education at scale.

We’re not regulating the questions: Who does the AI serve and how do we deal with conflicts of interest between child’s rights, family, school staff, the institution or State, and the company’s wants? Where do we draw the line between public interest, private interests, and who decides what are the best interests of each child?

We’re not managing what the implications are of the datafied child being mined and analysed in order to train companies’ AI. Is it ethical or desirable to use children’s behaviour as sources of business intelligence, to donate free labour in school systems performed for companies to profit from, without any choice (see UNCRC Art 32)?

We’re barely aware as parents, if a company will decide how a child is tested in a certain way, asked certain questions about their mental health, given nudges to ‘improve’ their performance or mood.  It’s not a question of ‘is it in the best interests of a child’, but rather, who designs it and can schools assess compatibility with a child’s fundamental rights and freedoms to develop free from interference?

It’s not about protection of ‘the data’ although data protection should be about the protection of the person, not only enabling data flows for business.

It’s about protection from strangers engineering a child’s development in closed systems.

It is about child protection from unknown and unlimited number of persons interfering with who they will become.

Today’s laws and debate are too often about regulating someone else’s opinion; how it should be done, not if it should be done at all.

It is rare we read any challenge of the ‘inevitability’ of AI [in education] narrative.

Who do I ask my top two questions on AI in education:
(a) who gets and grants permission to shape my developing child, and
(b) what happens to the duty of care in loco parentis as schools outsource authority to an algorithm?


UNCRC

Article 8

1. States Parties undertake to respect the right of the child to preserve his or her identity, including nationality, name and family relations as recognised by law without unlawful interference.

Article 18

1. States Parties shall use their best efforts to ensure recognition of the principle that both parents have common responsibilities for the upbringing and development of the child. Parents or, as the case may be, legal guardians, have the primary responsibility for the upbringing and development of the child. The best interests of the child will be their basic concern.

Article 29

1. States Parties agree that the education of the child shall be directed to:

(a) The development of the child’s personality, talents and mental and physical abilities to their fullest potential;

(c) The development of respect for the child’s parents, his or her own cultural identity, language and values, for the national values of the country in which the child is living, the country from which he or she may originate, and for civilizations different from his or her own;

Article 30

In those States in which ethnic, religious or linguistic minorities or persons of indigenous origin exist, a child belonging to such a minority or who is indigenous shall not be denied the right, in community with other members of his or her group, to enjoy his or her own culture

 

Data-Driven Responses to COVID-19: Lessons Learned OMDDAC event

A slightly longer version of a talk I gave at the launch event of the OMDDAC Data-Driven Responses to COVID-19: Lessons Learned report on October 13, 2021. I was asked to respond to the findings presented on Young People, Covid-19 and Data-Driven Decision-Making by Dr Claire Bessant at Northumbria Law School.

[ ] indicates text I omitted for reasons of time, on the day.

Their final report is now available to download from the website.

You can also watch the full event here. The part on young people presented by Claire and that I follow, is at the start.

—————————————————–

I’m really pleased to congratulate Claire and her colleagues today at OMDDAC and hope that policy makers will recognise the value of this work and it will influence change.

I will reiterate three things they found or included in their work.

  1. Young people want to be heard.
  2. Young people’s views on data and trust, include concerns about conflated data purposes

and

3. The concept of being, “data driven under COVID conditions”.

This OMDDAC work together with Investing in Children,  is very timely as a rapid response, but I think it is also important to set it in context, and recognize that some of its significance is that it reflects a continuum of similar findings over time, largely unaffected by the pandemic.

Claire’s work comprehensively backs up the consistent findings of over ten years of public engagement, including with young people.

The 2010 study with young people conducted by The Royal Academy of Engineering supported by three Research Councils and Wellcome, discussed attitudes towards the use of medical records and concluded: These questions and concerns must be addressed by policy makers, regulators, developers and engineers before progressing with the design, and implementation of record keeping systems and the linking of any databases.

In 2014, the House of Commons Science and Technology Committee in their report, Responsible Use of Data, said the Government has a clear responsibility to explain to the public how personal data is being used

The same Committee’s Big Data Dilemma 2015-16 report, (p9) concluded “data (some collected many years before and no longer with a clear consent trail) […] is unsatisfactory left unaddressed by Government and without a clear public-policy position.

Or see

2014, The Royal Statistical Society and Ipsos Mori work on the data trust deficit with lessons for policymakers, 2019  DotEveryone’s work on Public Attitudes or the 2020 The ICO Annual Track survey results.

There is also a growing body of literature to demonstrate what the implications are being a ‘data driven’ society, for the datafied child, as described by Deborah Lupton and Ben Williamson in their own research in 2017.

[This year our own work with young people, published in our report on data metaphors “the words we use in data policy”, found that young people want institutions to stop treating data about them as a commodity and start respecting data as extracts from the stories of their lives.]

The UK government and policy makers, are simply ignoring the inconvenient truth that legislation and governance frameworks such as the UN General Comment no 25 on Children in the Digital Environment, that exist today, demand people know what is done with data about them, and it must be applied to address children’s right to be heard and to enable them to exercise their data rights.

The public perceptions study within this new OMDDAC work, shows that it’s not only the views of children and young people that are being ignored, but adults too.

And perhaps it is worth reflecting here, that often people don’t tend to think about all this in terms of data rights and data protection, but rather human rights and protections for the human being from the use of data that gives other people power over our lives.

This project, found young people’s trust in use of their confidential personal data was affected by understanding who would use the data and why, and how people will be protected from prejudice and discrimination.

We could build easy-reporting mechanisms at public points of contact with state institutions; in education, in social care, in welfare and policing, to produce reports on demand of the information you hold about me and enable corrections. It would benefit institutions by having more accurate data, and make them more trustworthy if people can see here’s what you hold on me and here’s what you did with it.

Instead, we’re going in the opposite direction. New government proposals suggest making that process harder, by charging for Subject Access Requests.

This research shows that current policy is not what young people want. People want the ability to choose between granular levels of control in the data that is being shared. They value having autonomy and control, knowing who will have access, maintaining records accuracy, how people will be kept informed of changes, who will maintain and regulate the database, data security, anonymisation, and to have their views listened to.

Young people also fear the power of data to speak for them, that the data about them are taken at face value, listened to by those in authority more than the child in their own voice.

What do these findings mean for public policy? Without respect for what people want; for the fundamental human rights and freedoms for all, there is no social license for data policies.

Whether it’s confidential GP records or the school census expansion in 2016, when public trust collapses so does your data collection.

Yet the government stubbornly refuses to learn and seems to believe it’s all a communications issue, a bit like the ‘Yes Minister’ English approach to foreigners when they don’t understand: just shout louder.

No, this research shows data policy failures are not fixed by, “communicate the benefits”.

Nor is it fixed by changing Data Protection law. As a comment in the report says, UK data protection law offers a “how-to” not a “don’t-do”.

Data protection law is designed to be enabling of data flows. But that can mean that when state data processing rightly often avoids using the lawful basis of consent in data protection terms, the data use is not consensual.

[For the sake of time, I didn’t include this thought in the next two paragraphs in the talk, but I think it is important to mention that in our own work we find that this contradiction is not lost on young people. — Against the backdrop of the efforts after the MeToo movement and lots said by Ministers in Education and at the DCMS about the Everyone’s Invited work earlier this year to champion consent in relationships, sex and health education (RSHE) curriculum; adults in authority keep saying consent matters, but don’t demonstrate it, and when it comes to data, use people’s data in ways they do not want.

The report picks up that young people, and disproportionately those communities that experience harm from authorities, mistrust data sharing with the police. This is now set against the backdrop of not only the recent, Wayne Couzens case, but a series of very public misuses of police power, including COVID powers.]

The data powers used, “Under COVID conditions” are now being used as a cover for the attack on data protections in the future. The DCMS consultation on changing UK Data Protection law, open until November 19th, suggests that similarly reduced protections on data distribution in the emergency, should become the norm. While DP law is written expressly to permit things that are out of the ordinary in extraordinary circumstances, they are limited in time. The government is proposing that some things that were found convenient to do under COVID, now become commonplace.

But it includes things such as removing Article 22 from the UK GDPR with its protections for people in processes involving automated decision making.

Young people were those who felt first hand the risks and harms of those processes in the summer of 2020, and the “mutant algorithm” is something this Observatory Report work also addressed in their research. Again, it found young people felt left out of those decisions about them despite being the group that would feel its negative effects.

[Data protection law may be enabling increased lawful data distribution across the public sector, but it is not offering people, including young people, the protections they expect of their human right to privacy. We are on a dangerous trajectory for public interest research and for society, if the “new direction” this government goes in, for data and digital policy and practice, goes against prevailing public attitudes and undermines fundamental human rights and freedoms.]

The risks and benefits of the power obtained from the use of admin data are felt disproportionately across different communities including children, who are not a one size fits all, homogenous group.

[While views across groups will differ — and we must be careful to understand any popular context at any point in time on a single issue and unconscious bias in and between groups — policy must recognise where there are consistent findings across this research with that which has gone before it. There are red lines about data re-uses, especially on conflated purposes using the same data once collected by different people, like commercial re-use or sharing (health) data with police.]

The golden thread that runs through time and across different sectors’ data use, are the legal frameworks underpinned by democratic mandates, that uphold our human rights.

I hope the powers-at-be in the DCMS consultation, and wider policy makers in data and digital policy, take this work seriously and not only listen, but act on its recommendations.

Thinking to some purpose