Data Policy Digest
SOMETHING STIRS!
Hello, and welcome to our 29th Data Policy Digest, bringing you all - well, some of - the latest data and AI policy developments.
You may remember that we put the Digest on pause while we seek sustainable funding for it. Those efforts are ongoing (and please let us know if you have any feedback). But given just how much has been going on, we thought it was worth putting out a slimmed-down version (no, really, it actually is) to help you catch up on what’s been happening on data and AI policy, in the UK parliament and government, at least.
We also have an event coming up that we wanted to tell you about. On Thursday 20 March, we’re holding an in-person unconference in London (travel assistance available), Engaging people on data and AI: giving people and communities a powerful say on public sector technologies. It’s for people (and we’re particularly keen to reach those working inside the public sector) who are interested in ensuring that public, community and worker voices are heard in decisions about data, digital and AI. And if you’ve not been to an unconference before, you get to set the agenda and have the conversations you’d like to have.
Come along if you’re already working to engage the public through participatory processes or would like to do so in the future. Come along if you’re working directly with people and communities who are concerned about public sector uses of data, digital and AI, and want to make their voices heard. Come along if you’re engaging with members of the public on topics other than data, digital and AI and want to bring those topics into your conversations.
This is part of our wider project on Giving communities a powerful say in public sector data and AI projects. We’ve been running an online community of practice as part of that - you can sign up for the next one on Wednesday 12 March, where we’ll be reflecting on where public involvement in data and AI is at and where it should go next with the convenors of some other key networks.
If there’s something we’ve missed, something you’re up to that you’d like us to include next time or you have any thoughts on how useful the Digest is or could be, please get in touch via gavin@gavinfreeguard.com. A reminder that Connected by Data is on Bluesky @connectedbydata.bsky.social, and Mastodon @ConnectedByData@social.coop. You can also catch up on previous Digests.
To keep in touch with the next steps for the Data Policy Digest, sign up here.
Contents
Data policy developments
Bills, bills, bills
Or, pieces of legislation related to data and AI
Data (Use and Access) Bill (DUAB): Since our last edition, the DUAB (doo-wab?) has completed its Lords stages, with report stage and third reading. It’s had its first big moment in the Commons - second reading, when MPs debate the substance of a Bill for the first time. It’s now entered committee stage - you can get your evidence in to the Bill committee (but hurry - the committee meets for the first time on 4 March and will be accepting evidence until 18 March at the latest - ‘You are strongly advised to submit your written evidence as soon as possible’).
If you’d like to get (back) up to speed… the Open Rights Group had a briefing for second reading, and held an online event… there are briefings from Big Brother Watch, Public Law Project, Defend Digital Me, the National Data Guardian, NI Human Rights Commission, and the Commons Library… and there’s a working group on the Bill as part of the Data and AI Civil Society Network.
Also worth noting that the government has announced it will use the Bill to ban sexually explicit deepfakes… and there’s a call for evidence on computer evidence - you may remember the current presumption that evidence from software is assumed to be correct, which the Post Office Horizon scandal has challenged, to put it mildly.
AI Bill: We’re still awaiting what was expected to be a frontier-focused Bill (we think that the government needs to set out other steps alongside it). The government still intends to bring forward legislation, as per an answer in the Lords, but it’s taking longer than expected - Politico reported that may owe something to events across the Atlantic and was now earmarked for ‘spring’, while the Guardian goes for ‘summer’ (we may yet get an entire set of Vivaldi violin concerti before it’s actually published).
You may remember Tory peer, Lord Holmes, previously brought a private members’ bill on AI - he’s just published a report building on it.
Cybersecurity and Resilience Bill: Coming in 2025. There’s a consultation on ransomware, and an NAO report, to keep you busy in the meantime. Or Robert De Niro getting to grips with cyber attacks, if you’d prefer.
Online Safety Act: As ever, check out the Online Safety Act Network and their newsletter for more.
**Digital Markets, Competition and Consumers Act: **‘Government ousts UK competition watchdog chair’, was the BBC headline in January. The development prompted questions in Commons and Lords (and a ministerial statement), and fears from a pretty diverse coalition about new chair, and former Amazon UK boss, Doug Gurr.
Public Authorities (Fraud, Error and Recovery) Bill: Cast your mind back to the Data Protection and Digital Information Bill. Remember the controversial ‘DWP clauses’ giving the department powers to access data on people’s bank accounts? And the Labour government’s decision not to include that in the DUAB but to bring something forward separately?
The Public Authorities (Fraud, Error and Recovery) Bill has been introduced into the Commons and is at committee stage (the bill committee is taking evidence, until 20 March AT THE LATEST). The Commons Library has a briefing… the ICO has updated its response… Big Brother Watch have launched a campaign… Policy in Practice have taken a look… and Disability News Service and the Big Issue have some summaries, including of second reading debate.
Oh, and while we’re on financial checks… Home Office contractor collected data on UK citizens while checking migrants’ finances, reports The Observer.
**Other Bills: **The Children’s Wellbeing and Schools Bill is awaiting a date for report stage (Defend Digital Me have an update on what’s happened so far)… so is the Property (Digital Assets etc) Bill… and the Product Regulation and Metrology Bill starts report on 26 February.
Private Member’s Bills: Remember - these are unlikely to pass without government support, but…
In the Commons, Friday 7 March brings the second reading of several relevant bills, including Protection of Children (Digital Safety and Data Protection) Bill (Josh MacAlister, Lab), Children Not in School (Registers, Support and Orders) Bill (Paul Holmes, Con) and Public Body Ethnicity Data (Inclusion of Jewish and Sikh Categories) Bill (Preet Kaur Gill, Lab)…
In the Lords, both Lord Clement-Jones’ (Lib Dem) Public Authority Algorithmic and Automated Decision-Making Systems Bill and Baroness Owen’s Non-Consensual Sexually Explicit Images and Videos (Offences) Bill had their third readings on 20 January.
Digital centre of the universe
Or, what the Department for Science, Innovation and Technology and other parts of government have been doing on data and AI
Another quiet one 🙃
I promised this would be short, so…
AI The government published the long-awaited AI Opportunities Action Plan (and its response) on 13 January. AI teacher tools were the focus of one DSIT press release, small businesses another, inward investment yet another, while Peter Kyle’s written statement gave a broader summary and said he’d ‘front up’ any problems with public sector AI roll out, which is certainly an approach to accountability. Report author Matt Clifford has been named the PM’s AI Opportunities adviser (and had an updated register of interests published). The PM himself wrote for the FT - Britain doesn’t need to walk a US or EU path on AI: Going our own way will unleash national renewal. Minderoo, the ODI, the LGA, techUK and Dan McQuillan were among those publishing responses, along with political sketch writer Rob Hutton.
You can also express interest in becoming an AI Growth Zone… there was another release of algorithmic transparency records, including several from local government… GDS published an AI Playbook for the UK Government… the government responded to a previous select committee report on the governance of AI… the AI Cyber Security Code of Practice was updated following a consultation (more here, plus a whole collection of docs)… there’s a new collection of AI guidance on GOV.UK… World-leading AI trial to tackle breast cancer launched… UK-backed AI companies to transform British cancer care and spark new drug breakthroughs… Fellowships launched to explore how AI could change the way scientists drive new discoveries… AI and satellites speed up planning approvals by tracking wild habitats across England… i.AI’s Engagement and Strategy Manager wrote for the Heywood Quarterly on why the policy profession must embrace AI…
And then, of course, we had the AI Action Summit in Paris (10-11 February), with London fringe. Ahead of the event, the first International AI Safety Report was published and some additional organisations signed up to the commitments from the AI Safety Summit in Seoul (and New Zealand have been added to the original Bletchley declaration). But the big UK government story was the UK, along with the US (VP Vance’s speech was another major talking point), not signing the Summit declaration (coverage from BBC, Guardian, Sky). Chi Onwurah, chair of the Science, Innovation and Technology select committee, made a statement and asked the government about it. (And while we’re on the Summit, we were delighted to be part of the Participatory AI Research & Practice Symposium - Tech Policy Press have a few essays on the event.)
And last but not least… on 14 February, the AI Safety Institute became the AI Security Institute - ‘Tackling AI security risks to unleash growth and deliver Plan for Change’ surely wins government buzzword shoehorn bingo - and the government signed an agreement with AI giant, Anthropic. That same day (not entirely coincidentally, one imagines), Peter Kyle addressed the Munich Security Conference (somewhat overshadowed by another Vance speech).
Digital centre Forget AI - the machinery that the really cool kids care about is machinery of government. On 21 January, the government announced that the Government Digital Service would expand to include the Central Digital and Data Office, Incubator for AI and the Geospatial Commission (and bits of the Responsible Technology Adoption Unit), alongside a review of the current state of digital government (previewed with ‘Archaic tech sees public sector miss £45 billion annual savings’), a wider blueprint for the future and other announcements including new apps for the public and a new suite of AI tools for public servants. And on the subject of rejigging the digital centre - the government chief data officer is departing after less than a year in post, while DSIT has disbanded the One Login Inclusion and Privacy Advisory Group as the digital centre thinks about its approach to external engagement.
The State of Digital Government Review touched on the data and digital initiatives in the government’s major projects portfolio, the latest report on which has just been published (DSIT data here, as details emerge on government’s new major projects body). Additionally, Algorithmic transparency requirement added to digital spend controls, reports Public Technology.
And DSIT also got in on the National Procurement Policy Statement fun (the zenith of stuff that sounds dull but is actually very important for how government works).
Copyrights and wrongs The government consultation on AI and copyright closed earlier this week. The Guardian reports that ‘UK ministers consider changing AI plans to protect creative industries’, with notable campaigns from the creative and media industries grabbing attention. The chairs of the Science, Innovation and Technology, and Culture, Media and Sport select committees have written to ministers, saying that ‘The principle that everyone should receive fair remuneration for their creative work should underpin the relationship between the AI sector and rights holders’.
Minderoo, the Bennett Institute and ai@cam at the University of Cambridge published a big report on the topic.
And worth keeping an eye on the EU situation and this court case, too.
Other DSIT news: The big one: after eleven years, the UK government has a new digital inclusion strategy (see also: press release). You can respond to the consultation (closes 9 April) or even express interest in joining the Digital Inclusion Action Committee (which we mentioned last time). The Good Things Foundation have been pushing for a plan for a while…
Baroness Bertin has submitted her review on online pornography to the government, who have responded (the BBC have a summary). It was commissioned by the previous government back in December 2023, so it’s been a long time… being prepared.
The planned Manchester Digital Campus, which ‘will bring together a number of Civil Service departments with a focus around digital skills’, has been given a green light… while the IfG’s Whitehall Monitor took a deep dive into DSIT’s morale…
And… The government published its response to a consultation on proposed changes to the data protection fee regime… Colin Hughes appointed Chair of Ordnance Survey… Global Coalition on Telecommunications: principles on AI adoption in the telecommunications industry… UK-Japan Digital Partnership… UK Science and Innovation Network summaries for Singapore, UAE, Switzerland… UK redoubles Horizon push as Kyle forges deeper R&I links with EU… minister Lord Vallance added Oxford-Cambridge Innovation Champion to his portfolio of roles… minister Baroness Gustafsson weighed in on business people being attracted to government… International Coalition on Cyber Security Workforces… UK government seeks out quantum industry experts for advisory board to accelerate deployment of game-changing technology… Platform design and the risk of online violence against women and girls… DSIT evaluation strategy… Smart Machines Strategy 2035…
And finally… on the National Data Library: Wellcome and the ESRC published the submissions to their technical white paper challenge (the Bennett Institute - the Oxford one - have blogged about theirs, as have the ODI), TBI/The Entrepreneurs Network published a report, and Politico notes that a ‘reboot’ of sorts is planned under new leadership - Sarah Hodgetts is now listed on GOV.UK as ‘Director of the National Data Library and Geospatial, as part of the Government Digital Service’.
Other bits of government: Let’s start with foreign policy, shall we? A week ago - it feels like a lot longer - The Sunday Times reported that Downing Street was keen to get closer to the US on AI. Politico suggested similar. Last Friday morning, health sec Wes Streeting said the NHS would not be on the table in a trade deal with the US, but suggested health data might be. Then, of course, the world changed.
While we’re on the US… it seems the US didn’t know about the UK government’s demand to access encrypted Apple data which led to the company pulling its Advanced Data Protection tool in the UK. Turns out tech companies may respond to hard power and not just ‘humility’ after all. (ORG have launched a very timely campaign on saving encryption.)
Next, the traditional ‘centre’ - No 10, Cabinet Office and Treasury. There’s a big spending review coming - Lord Vallance told parliament that it will upgrade ageing tech and fund AI, while chief secretary to the Treasury Darren Jones covered AI and much else in a speech and questions at the IfG. And Jeni pondered the challenge of saving money with AI in the UK public sector in the spending review context.
On DWP… there are reports the right to switch off is among workers’ rights reforms to be dropped (Times)… AI prototypes for UK welfare system dropped as officials lament ‘false starts’ (Guardian)… Liz Kendall confirms government will no longer back controversial ‘computer says yes’ system (Guardian)…
Someone at the Home Office is briefing that research on public attitudes towards facial recognition tech is a waste of money since systems are already in use (Times)… the head of NHS England is stepping down (BBC)… Labour will use AI to snoop on social media (Telegraph)… [Northern Ireland] Minister sidesteps question about NICS jobs risk from artificial intelligence (CSW)… the NAO has a report on the government’s approach to technology suppliers…
And former tech ministers, Matts Hancock and Warman, have shared some of their reflections on being in government, while Damian Collins has started a Substack.
Parly-vous data?
Or, what’s been going on in parliament
Previous Commons business… Mon 13 Jan was busy, with a ministerial statement (plus contributions from others) on the AI Opportunities Action Plan, statement from the defence select committee on their ‘Developing AI Capacity and Expertise in UK Defence’ report, and a Westminster Hall debate on Children’s Social Media Accounts… Tue 14 Jan, the Science, Innovation and Technology Committee hosted an evidence session on Innovation, growth and the regions… Fri 17 Jan, second reading of private members’ bills, including the Social Media (Access to Accounts) Bill (Max Wilkinson, Lib Dem)… Tue 21 Jan, the Science, Innovation and Technology committee took oral evidence in its inquiry into social media, misinformation and harmful algorithms… Thu 23 Jan, where subjects at Cabinet Office questions included cybersecurity, and the Public Accounts Committee held an evidence session on Whole of Government Accounts… Mon 27 Jan, AI and copyright was one of the topics raised in a debate on the creative industries… Tue 28 Jan, the Science, Innovation and Technology committee took evidence on the role of science advice in government… Thu 30 Jan, the DSIT and Cabinet Office perm secs and the government chief technology officer gave evidence at the first hearing of PAC’s ‘use of AI in government’ inquiry…
Tue 4 Feb, the Culture, Media and Sport committee took evidence on AI and copyright and the Treasury committee from the ONS on economic statistics… Tue 11 Feb, the Science, Innovation and Technology committee quizzed science minister, Lord Vallance, while the Foreign Affairs Committee held the first session of its inquiry on disinformation diplomacy… Wed 12 Feb was DSIT questions, and the second reading of the Data (Use and Access) Bill (see above)… Thurs 13 Feb brought a written ministerial statement on the National Procurement Policy Statement which includes a bit on AI… Mon 24 February, Tony Vaughan (Lab) had a Westminster Hall debate on an e-petition relating to a minimum age for social media (the Commons Library has a debate pack)… it was also Home Office questions, where one of relevant issues raised was Apple and encryption… Tues 25 February, the Science, Innovation and Technology committee quizzed Google, and then Meta, TikTok and X as part of its Social media, misinformation and harmful algorithms inquiry, while PACAC quizzed the new Cabinet Secretary… Wed 26 February, former DCMS Secretary Jeremy Wright (Con) had a Westminster Hall debate on Implementation of the Online Safety Act… Thurs 27 February brought DCMS questions (copyright included), and the Public Accounts Committee quizzed senior civil servants on Government’s relationship with digital technology suppliers…
Coming up in the Commons… Thurs 6 March, Cabinet Office questions… Fri 7 March, the private members’ bills up for second reading include Protection of Children (Digital Safety and Data Protection) Bill (Josh MacAlister, Lab), Children Not in School (Registers, Support and Orders) Bill (Paul Holmes, Con) and Public Body Ethnicity Data (Inclusion of Jewish and Sikh Categories) Bill (Preet Kaur Gill, Lab)… Mon 10 March, the PAC takes evidence from senior civil servants on Government cyber resilience… Wed 26 March, DSIT questions…
Previous Lords business… On Tues 14 Jan, the European Affairs Committee wrote to Peter Kyle about data adequacy with the EU… Thurs 16 Jan, a question from Baroness Ritchie of Downpatrick (Lab) on ‘Risks posed by small online platforms‘ and a statement from Lord Vallance on the AI Opportunities Action Plan… Mon 20 Jan, committee stage for private members’ bills including Baroness Owen’s (Con) Non-Consensual Sexually Explicit Images and Videos (Offences) Bill, and Lord Clement-Jones’ (Lib Dem) Public Authority Algorithmic and Automated Decision-Making Systems Bill… Tue 21 Jan, the Data (Use and Access) Bill started report stage (see above)… Mon 27 Jan, there was a question from Viscount Stansgate (Lab) on ‘Energy usage involved in mining Bitcoin and in general internet activity’… Thu 30 Jan, there was a short debate prompted by Lord Foster (Lib Dem) on the creative industries, AI and copyright…
Fri 7 Feb, it was the third reading of the Owen (explicit images and videos) and Clement-Jones (algorithms) bills… Mon 10 Feb, Lord Holmes (Con) asked about plans to regulate AI… Wed 12 Feb, Lord Kamall (Con) had a question on Electronic Patient Record Systems in the NHS… Mon 24 February, there were a couple of motions on regulations under the Online Safety Act… and Thurs 27 February, Baroness Chakrabarti (Lab) had a question on AI and copyright, while Lord Beamish (Lab) had a debate on Post Office Horizon compensation…
Coming up in the Lords… Thurs 6 March, government minister Baroness Gustafsson (Lab) leads a debate on women’s participation and leadership in science and technology for International Women’s Day…
And the latest list of all-party parliamentary groups (informal groups of MPs and peers interested in specific subjects) is available.
In brief
Or, a quick overview of some other key developments
States of play: Well. It’s all going very well, isn’t it? Threw was an early executive order on ‘removing barriers’ to US leadership on AI, and then… Birkbeck’s Ben Worthy has blogged about how Donald Trump is plunging government websites into darkness, with profound consequences for American democracy, while Tech Policy Press have several pieces on DOGE’s activities, as does Wired and 404 Media, while 538 also look at data deletion and Politico’s West Wing Playbook looks at DOGE and data. The NYT has some detailed reporting on the Musk takeover of the federal bureaucracy - the Atlantic made an earlier attempt.
And there’s also a statement from staff at the recently shuttered digital agency, 18F.
Here’s looking at EU: The new Commission Work Programme (for the next year) has a fair few data and digital bits and pieces, although not an expected AI Liability Directive… and in Germany, Politico have taken a look at How Germany’s far right is harnessing AI to win votes (they won a fair few).
Thinktanks and civil society: At Connected by Data, we added some case studies to our community campaigners toolkit; contributed to the development of pioneering new standards on the use of AI to manage workers in Wales; Jeni and I contributed to TPX Impact’s State of Data report; and we worked with the Ada Lovelace Institute and Demos on what civil society and others would like to see alongside a frontier AI Bill…
Speaking of whom… Demos have been very busy with reports on epistemic security and information supply chains, election law in the age of AI, and advancing digital rights, and coordinating an open letter on the need for a citizen-led declaration of digital rights and principles… and Ada with previewing the AI Action Summit, a report on AI and education (they’ve blogged on that, and AI companions and access to data frameworks), and another on public compute…
I think I managed to miss the ODI’s new AI Data Transparency Index - they’ve also published new reports on topics including Privacy Enhancing Technologies, and various blogposts, while I was one of the contributors to a policy webinar looking ahead at data and AI policy in the UK, Europe and US in 2025…
The Institute for the Future of Work published their major Pissarides Review into the Future of Work and Wellbeing (launching it with a conference), and reports on how tech adoption is reshaping work and looking at the distribution of capabilities amidst technological change…
JRF published their reflections on a workshop about AI and countervailing power in the UK, and gave an update on their AI work…
Careful Industries had a new report on inclusive innovation, while Rachel Coldicutt blogged for the LSE on the winners and losers from the AI Summit…
Privacy International, the TUC and others called for algorithmic transparency from delivery apps… ORG published a report on electoral canvassing apps and data privacy… Justice published a report on AI in the justice system… AWO won a landmark case under GDPR… the Health Foundation published Tech to save time: how the NHS can realise the benefits… while both IfG and Reform (the thinktank, etc) had reports on AI and government.
In Labour land… IPPR had a new report on The new politics of AI: Why fast technological change requires bold policy targets… Labour Together wrote about Digital ID (something which Labour MPs have been talking more and more about, as has, well of course) and reacted to the AI Opportunities Plan… while the TBI have new reports or commentary on the National Data Library (with The Entrepreneurs’ Network), Accelerating UK Science in the Age of AI, smart data, DOGE, the AI Action Plan and How Leaders in the Global South Can Devise AI Regulation That Enables Innovation… and the Social Market Foundation had a report on How to power AI: Boosting compute capacity for UK AI.
Coming up: I’ll be hosting the next couple of Data Bites events on 5 March (speakers include i.AI, the HFEA and Full Fact) and 3 April…
On 14 March, Careful Trouble are kicking off a reading group for hopeful technologists (and pondering something more)…
Simon Cross (Church of England) is holding a Data and AI Civil Society Network event in Oxford, AI and Civil Society WHO/HOW?, on 3 April…
And a happy Open Data Day (/week) to you all!
Good reads
Or, some other bits and pieces
- Walgreens’ Fridge Fight Bodes Poorly for Future of Retail (Bloomberg)
- Candy Crush, Tinder, MyFitnessPal: See the Thousands of Apps Hijacked to Spy on Your Location (404 Media)
- FTC Takes Action Against General Motors for Sharing Drivers’ Precise Location, Driving Behavior Data (Federal Trade Commission)
- She is in love with ChatGPT (New York Times)
- I doorknocked for Labour, then racist deepfake ruined my life (The Times)
- ‘Amateur and dangerous’: Historians weigh in on viral AI history videos (BBC)
- Life Is More Than an Engineering Problem (Los Angeles Review of Books)
- Spy vs. AI (Foreign Affairs)
And finally…