Data Policy Digest

Gavin Freeguard

Gavin Freeguard

Hello, and welcome to our seventeenth Data Policy Digest, bringing you all the latest data and AI policy developments.

And, well. Yes. Quite.

June is bustin’ out all over and we’re all aboard the dizzying electoral carousel. Work plans have been torn up, holidays cancelled, and newsletter drafts repurposed as the UK hurtles more quickly than most of us expected towards a general election.

You can expect this newsletter to be more frequent and - most importantly - shorter through the campaign. (Dear god, let it be shorter - the campaign feels long enough already.)

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@connectedbydata.org. We’re on Twitter @ConnectedByData and @DataReform. You can also catch up on previous Digests.

To receive the next edition of the Data Policy Digest direct to your inbox sign up here.

Contents

Data policy developments

What we’ve been up to

What everyone else has been up to

Events

Good reads

Data policy developments

Deeply DPDIB

The Data Protection and Digital Information Bill has fallen.

Not the latest and nerdiest entry in the Gerard Butler/Morgan Freeman film franchise, but a significant moment for us data policy folk. To put it more straightforwardly (but equally cinematically): Ding-dong, the Bill is dead.

Rishi Sunak’s decision to dissolve parliament and call an election means we entered ‘wash-up’, defined by parliament as ‘the last few days of a Parliament before dissolution. Any unfinished business is lost at dissolution and the Government may need the co-operation of the Opposition in passing legislation that is still in progress.’ (Not to be confused with ‘wash-out’, defined by me as an historically bad piece of political comms announcing the election which may yet turn out to be a pathetic fallacy for the election result. You can insert your own ‘Things Can Only Get Wetter’, ‘Tory wet’ and ‘Drowning Street’ jokes.)

Some Bills were rushed through. Some, including DPDIB, were not.

The Open Rights Group and Big Brother Watch were among those hoping it wouldn’t pass; the Digital Marketing Association and techUK were among those hoping it would. Politico reported that ‘the government wouldn’t budge over late amendments giving officials powers to inspect bank accounts of benefit recipients’, even though the government had apparently been considering some concessions to get the Bill through even before the election was called. Mishcon de Reya’s Jon Baines and UKAuthority were among those with reaction.

So the UK’s data regime stays as before - for now. Those DWP powers are gone, though they weren’t (as some suggest) the only things wrong with a missed opportunity of a bill that weakened safeguards and swept away rights. Less controversial changes have also been put on ice, including provisions on smart data, the National Underground Asset Register, digital identity services, and converting the Information Commissioner’s Office (a ‘corporation sole’ in the legal jargon) into the Information Commission - the ICO had just published ads for an Interim Chief Executive Officer - Fixed Term Contract (up to two years) and some non-executive directors in expectation (follow the links and you’ll see both have disappeared). The Biometrics and Surveillance Camera Commissioner survives, when it had been expected to be a transitional appointment. There might also have been something coming on researcher access to tech company data - at DSIT questions, minister Saqib Bhatti told a questioning Miriam Cates to ‘watch this space’. The Camrose/Kidron amendment allowing coroners to require information from tech companies regardless of the cause of a child’s death has also fallen. (An issue unlikely to go away - today The Guardian has a story, UK mother of boy who killed himself seeks right to access his social media.)

Politico reports that Labour could ‘revive parts of the Bill’ if it wins the election: ‘one person close to the Labour frontbench said the party is keen on some aspects of the bill, such as proposals to broaden data use for research purposes — while there have also been suggestions the party could wrap up aspects of the data reform within a wider AI bill’.

For completeness… our two previous Digests summarised Lords committee stage, the most consequential 7 days since Craig David… pre-election announcement, dates had been given for Lords report stage: Monday 10 June and Wednesday 12 June, which is when the Lords would have had another go at debating, scrutinising and amending the Bill (including being able to vote on the latter)… some ‘will write’ letters, from ministers to peers, have been added to the Bill page…

One of the subjects covered was those DWP powers - to access data about bank accounts of benefit claimants (including pensions) and those associated with them - which continued to generate most controversy and coverage… Big Brother Watch delivered a bumper petition to Downing Street (Petition against DWP’s plans for bank ‘mass spying’ powers nears 200k signatures, reported City AM, who also had an editorial on how Proposed snooping powers are an incursion into our most basic rights)… new research from Demos - “This system is rigged”: A social media listening exploration revealing how people are talking about poverty online - found concerns are cutting through to the public (see page 40)… DWP anti-fraud measures will allow monitoring of bank accounts of landlords, carers and parents (Computer Weekly)… DWP sparks outcry with plan to hire 2,500 ‘agents’ to investigate universal credit claimants (Big Issue)… MPs had tabled an early day motion, Government surveillance of bank accounts (‘Another Horizon’ – MPs table motion against DWP bank-monitoring proposals is the Public Technology headline)… ‘This measure is expected to save the taxpayer £600 million over five years’ says an updated fraud plan from DWP… while various outlets (eg the Birmingham Mail) wrote up relevant bits of the impact assessment into a story…

This all came around further DWP/data concern, with Schools Week reporting a DfE/DWP link - Revealed: Secret deal to let benefit fraud squad snoop on pupil data - thanks to work from Defend Digital Me…

Additionally… openDemocracy looked at the state of Subject Access Requests even before the Bill’s planned changes to them, Revealed: ‘Wild West’ for personal data undermines UK human rights… anyone interested in the democratic engagement clauses might like new academic research, Citizens’ Acceptance of Data-Driven Political Campaigning: A 25-Country Cross-National Vignette Study (Social Science Computer Review), this piece on how Campaign microtargeting and AI can jeopardize democracy (LSE) and be interested in the Tories self-refer[ring] to ICO over data breach (Computing)… on international transfers, It’s time to ban TikTok for the sake of our democracy and security, wrote Lord Alton, in support of Lord Bethell’s amendment, for PoliticsHome… on algorithmic transparency, we were told to expect a set of releases by central government departments by the end of July… the Law Society Gazette reports on some controversy around home buying - one of the Smart Data sectors likely to be supported by the Bill…

And in other news, Computer Weekly asks What does the Data Protection and Digital Information Bill mean for UK industry?… the Welsh Government has asked for discussions to ‘facilitate progress’ on the Bill… The UK’s quest to boost digital surveillance in 2024 isn’t over, is TechRadar’s take… I’m nearly a year late to this, but Andrew Harvey, of the National Health and Care Strategic Information Governance Network, wrote about the Likely Implications of the Data Protection and Digital Information Bill on the Health and Care Sector… and while we’re on data protection, keep an eye on Schrems III: Caped hero takes on giant - again (Law Society Gazette).

DPDIB: gone, but not forgotten. Congratulations to everyone who contributed to scrutinising and ultimately stopping the proposed damage to our data protection regime. We, and the data and AI civil society network, will continue to work on how that regime should be improved, whoever wins the election.

Bills, bills, bills

Digital Markets, Competition and Consumers Act 2024 This one did make it through and has received Royal Assent to become law. The Competition and Markets Authority immediately published a Consultation on digital markets competition regime guidance, open until 12 July. (They also have an open consultation on ‘the proposed variation to the commitments in respect of Meta’s use of advertising data’, while their Microsoft/Mistral AI partnership merger inquiry found it did not qualify for investigation under the Enterprise Act.)

Online Safety Act ‘Nearly 1400 pages. FOURTEEN HUNDRED.’

No, not the latest Digest, but Ofcom’s proposed measures to improve children’s online safety… Consultation: Protecting children from harms online (Ofcom)… Tech firms told to hide ‘toxic’ content from children (BBC)… ‘We owe our children’: Grieving parents say social media costs lives (BBC)… Can we really ‘reset the internet’ to make it safer for children? (BBC)… Could Ofcom ban social media for under-18s? (BBC)… ONLINE SAFETY PROPOSALS COULD CAUSE NEW HARMS WARNS OPEN RIGHTS GROUP (ORG)… explainer on the Online Safety Act (DSIT)… and guidance to Ofcom on determining fees payable by regulated services under the Act…

Something for Haidters, Disconnect: The Case for a Smartphone Ban in Schools (Policy Exchange)… and haters, Do smartphones really cause mental illness among adolescents? Ten problems with Jonathan Haidt’s book (LSE)… Is there a middle way on children and smartphones? This researcher thinks so (the FT interviews Sonia Livingstone)… THE PANIC OVER SMARTPHONES DOESN’T HELP TEENS (The Atlantic)…

And… Ofcom investigates OnlyFans’ age verification measures (Ofcom)… Ofcom accused of ‘excluding’ bereaved parents from online safety consultation (The Observer)… Australia’s online safety regulator has drawn a line in the sand for X. Will she prevail? (The Guardian)… elsewhere, ICCL expresses dismay as ‘toxic’ algorithms not covered by online safety code (RTE)… and Kids Off Social Media Act Would Harm Childrens’ Safety and Privacy Online, Say Civil Society Organizations (New America).

And in parliament… there was a Westminster Hall debate in parliament on Smartphones and Social Media: Children (Miriam Cates MP, 14 May)… and MPs urge under-16s UK smartphone ban and statutory ban in schools, writes the Guardian about a pre-dissolution report from the Education Select Committee.

Other The Automated Vehicles Bill is now the Automated Vehicles Act, as of Monday 20 May - the Commons Library has a briefing while there’s been a lot of attention on some of the private sector players, eg Wayve’s charm offensive: How slick lobbying helped bag Europe’s biggest AI deal (Politico), and Vote of confidence in UK economy as British AI company Wayve secures over $1 billion to develop AI for self-driving vehicles (DSIT)… there were several relevant (and unsuccessful) amendments to the Criminal Justice Bill, including Sarah Owen (Lab) wanting to make AI election misinformation illegal, and Caroline Nokes (Con) on non-consensual intimate images… and the Artificial Intelligence (Regulation) Bill, Lord Holmes’ private members’ bill, has also fallen with the prorogation of parliament.

Never mind the ballots

News Tory ‘China is looking to dominate the 21st century by stealing a lead in technology’ was the only mention of tech or data in the PM’s speech announcing the election. AI got more of an outing at his speech on security the previous week, which acknowledged heightened fears about AI. ‘But the paradox of our age is that for all the profound dangers we face, right now, we also hold in our hands an opportunity for human progress that could surpass the industrial revolution in speed and breadth. Technologies like AI will do for the 21st century what the steam engine and electricity did for the 19th.’ He also spoke to the I about Revolutionary AI trial offers breast cancer hope for millions on NHS, Sunak reveals around the Seoul AI Summit, which the UK co-hosted.

The main Tory/data story so far is an unexpected one: Rishi ran Peloton 5k in California after calling election (Guido Fawkes). Labour also had some fun with the confluence of rumours about Sunak’s next steps post-election and Tory attacks on ‘mickey mouse degrees’ with an unfortunate photo.

Labour movement Keir Starmer’s speech upon the election being called railed against the current way of doing politics ‘that is at best doing something to people, not with them. But at its worst, as we saw in horrifying detail in Westminster last week, those twin injustices – the Horizon and Infected Blood scandals, is something much, much darker even than that.’

AI and data may not end up as a major campaign issue in its own right - Starmer’s first major speech didn’t mention it - but it may be integral to policy and delivery in other areas and could be a dividing line between the parties. We saw a bit of that - on AI regulation - in the debate following a ministerial statement on the AI Seoul Summit. Labour underlined that it would use legislation to force AI companies on safety testing, in contrast to the Conservatives’ voluntary approach, and criticised the government for not settling the copyright and IP questions arising from AI.

The latest version of Labour’s New Deal for Working People - ‘Labour’s Plan to Make Work Pay’ - has a few things on tech, including a Labour government working with workers, unions, employers and experts ‘to examine what AI and new technologies mean for work, jobs and skills, and how to promote best practice in safeguarding against the invasion of privacy through surveillance technology, spyware and discriminatory algorithmic decision making’ - ‘at a minimum [ensuring] proposals to introduce surveillance technologies would be subject to consultation and negotiation’. They would also ‘update regulations to outlaw the use of predictive technologies for blacklisting and safeguard against singling out workers for mistreatment or the sack without any evidence of human interaction’ and introduce a right to switch off.

Also… the list of business leaders backing Labour ahead of Rachel Reeves’ first big speech includes some from the world of tech… Labour, the vampire kangaroo and a very ‘spooky’ dinner engagement is openDemocracy’s take on some lobbying they’ve been investigating… there would be a ‘rapid 100-day security review to identify all major threats facing Britain from day one of government’, including ‘generative AI and cyber hacks’… there will be a new infrastructure body… Politico’s Morning Tech UK reports that Faculty - co-authors of a recent report on AI and public services with the Tony Blair Institute - have joined Public Digital in seconding a staff member to Peter Kyle’s office… while Labour Digital has published a list of PPCs it is endorsing to ‘harness the power of technology to empower people, transform society and grow our economy’ (first 20 trailed on LinkedIn, full 40 on X-formerly-known-as-Twitter)… a former Labour adviser gives advice in Talking tech: Labour and the startup ecosystem (Headland)… MPs’ summer break could be cut short if Labour wins as Reeves eyes early Budget (the I), while there may be a battle for jobs (not least because there are more shadow frontbenchers than there are paid ministerial posts)… and an unexpected story with a data protection angle, as Labour candidate told police not to respond in row (BBC).

Before all that happened (i’d already drafted it, so you’re going to have to read it - or skip over it, up to you)… Keir Starmer had been due to give a big speech on tech at a Labour Tech Summit on 3 June - now cancelled, with Politico also reporting an AI strategy reveal would have come at London Tech Week on 10 June… shadow leader of the House (and former shadow DCMS secretary) Lucy Powell gave a keynote speech on improving the legislative process (something we’ve come to know far too well in recent months - and yes, familiarity has bred contempt), following an interview with the I… Westminster-watching journalist Ian Dunt gave a useful summary with his verdict…

Lots of pieces on Labour’s mission-oriented approach to government and a little controversy over attribution… What Mission-Driven Government Means (Mariana Mazzucato and Rainer Kattel)… Mazzucato is one of the authors and speakers at a report launch (Thursday 30 May - said report published just as we went to pixel)… Keir Starmer’s real mission: to overhaul how government works (Andrew Marr, New Statesman)… and PPC Jake Richards wrote about How embracing technology can restore confidence in the police - part of one of the missions - for the Social Market Foundation…

There were also various pieces on ideas and internal influences, such as Why Labour staffers and MPs don’t say no to ‘power behind the throne’ Sue Gray (Guardian - 20 bills are apparently in development)… What is Starmerism? (New Statesman)… The McSweeney Project (UnHerd)… Morgan McSweeney, the workaholic Irishman who built Starmer’s Labour (Sunday Times)… including from or about Labour Together, such as Starmerism is not at war with Blairism (‘On technology, Labour is optimistic about its potential to improve public services. But it is sceptical that technology is neutral and conscious of the need to proactively and intentionally shape how and where it’s adopted’ - newly-selected Josh Simons for the New Statesman)… The think-tank laying the groundwork for a Labour government (FT on Labour Together)… What does Labour Together want? (New Statesman).

Meanwhile, in Wales, first minister Vaughan Gething will face a no confidence vote this week, after sacking a minister and Plaid Cymru pulled out of a cooperation deal (BBC).

A farewell to harms The first tech issue to properly break through into the campaign so far would seem to be online safety. Today, the Molly Rose Foundation has published a five point plan, which includes a new Online Safety Act, greater tech company transparency and a one-off windfall tax. ‘The tech giants must be more open’ is the Guardian’s editorial take. Yesterday, shadow DSIT secretary Peter Kyle confirmed Labour would introduce legislation on coroner access to children’s data after a child’s death. Labour’s Yvette Cooper and Conservative Victoria Atkins were quizzed about access to children’s data on the BBC’s Sunday with Laura Kuenssberg (though the question seemed to be about parental access, rather than coroner access). ‘Labour promises to give bereaved families access to children’s social media “as fast as possible”’ is the Independent’s take on that discussion (they’ve also clipped the relevant questions).

Everything else Rishi Sunak isn’t the only party leader to have gotten soaked so far - Lib Dem leader Ed Davey did too. (You can insert your own ‘Swimming Here’, ‘Lib Dems buoyant’ or ‘floating voter’ jokes.) Tax online giants to help kids’ mental health, say Lib Dems is their main digital/data line so far.

We’ll be keeping an eye out for manifestos from civil society and others during the campaign - get in touch if you’ve spotted or are publishing something. As well as the civil society Progressive Vision (pamphlet version) and the recently-launched Open Data Institute manifesto, there’s also… OpenUK’s Open Manifesto 2024… Seven exponential policies for a new government (Azeem Azhar, Exponential View)… techUK’s general election hub… Full Fact’s election live blog… Fix the digital divide – for good from Good Things Foundation… free refurbished digital devices and training for the digitally excluded, and automatic social tariffs through data matching can be found in the Fabian manifesto… Charity sector bodies call on next government to establish ‘civil society charter’ (Civil Society)… and The British Academy calls for next government to strengthen humanities and social sciences to tackle the UK’s biggest challenges. Their manifesto mentions the need for new data infrastructure but also for SHAPE subjects (social sciences, humanities and the arts) to be brought to questions on uses of technology.

More officially, there’s the Cabinet Office Election guidance for civil servants, while the Office for Statistics Regulation also has a hub up and running with guidance.

And… UK political parties urged to commit to AI honesty before elections (Computing)… Is AI set to destroy trust in elections? Tackling misinformation in politics & Parliament (Hansard Society with Full Fact’s Chris Morris)… AI fakes, abuse and misinformation pushed to young voters on TikTok (BBC)… Is this the TikTok election? (The Media Show, BBC)… YouGov polled deepfake worries… ‘Show this to everyone’: UK political ads move away from microtargeting (The Guardian)…

And finally… the FT have a prediction/model tool alongside their poll tracker… Public First are doing God’s work with this policy tracker… and here, have some bar chart beef.

AI got ‘rithm

Let’s try to keep this short. (And fail, miserably.)

Summit going on The UK co-hosted the AI Seoul Summit with South Korea, the summit title newly shorn of ‘safety’. UK Rishi Sunak co-authored a piece with South Korean president Yoon Suk Yeol (his Seoul-mate? So he wasn’t the Seoul author? geddit?!?!), Only global AI standards can stop a race to the bottom (the I). Michelle Donelan took questions on LinkedIn.

This is what DSIT published about it… AI Seoul Summit programme… New commitment to deepen work on severe AI risks concludes AI Seoul Summit… Seoul Ministerial Statement for advancing AI safety, innovation and inclusivity: AI Seoul Summit 2024… Secretary of State Michelle Donelan opens the AI Seoul Summit… Frontier AI Safety Commitments, AI Seoul Summit 2024… Seoul Declaration for safe, innovative and inclusive AI: AI Seoul Summit 2024… it followed publication of the first International Scientific Report on the Safety of Advanced AI (press release - Report highlights disagreement among experts on AI safety was Computer Weekly’s take)… and was accompanied by various other international announcements from the UK, including Historic first as companies spanning North America, Asia, Europe and Middle East agree safety commitments on development of AI…

You may remember when the UK hosted the first AI (Safety) Summit, we in UK civil society were Not Happy at our voices not being heard. Sadly our Korean counterparts had similar complaints. We also put out a statement - with the TUC and ORG, co-organisers of our open letter from the autumn:

The Summit in Seoul looks to be a departure from the X-risk hype that engulfed the UK’s summit in 2023. There is a welcome focus on the social, economic and environmental implications of AI that civil society worldwide has been demanding. However, we’re yet again seeing civil society’s voice sidelined in favour of corporations and governments. This lesson from the U.K. Summit does not appear to have been learnt. Let’s hope France in 2025 does better!

The Guardian had a couple of summary pieces - Trying to tame AI: Seoul summit flags hurdles to regulation, and What we learned from the global AI summit in South Korea - and spoke to Max Tegmark, Big tech has distracted world from existential risk of AI, says top scientist.

There were also AI Safety Institute developments around the Summit… Government’s trailblazing Institute for AI Safety to open doors in San Francisco, a UK-Canada science of AI safety partnership, Global leaders agree to launch first international network of AI Safety Institutes to boost cooperation of AI, AI Safety Institute releases new AI safety evaluations platform (all DSIT - see also new research funding and grants)… Ada had a piece, Safety first?, Reimagining the role of the UK AI Safety Institute in a wider UK governance framework… while the US AI Safety Institute published its strategic vision…

And… as France picks up the baton, Politico says it’ll be February for the AI Opportunities Summit… while 5 June brings an AI Fringe event at the British Library.

International If you’re looking for Summit else international… In Europe… Artificial intelligence (AI) act: Council gives final green light to the first worldwide rules on AI (European Council)… Commission establishes AI Office to strengthen EU leadership in safe and trustworthy Artificial Intelligence (European Commission)… A vision for the AI Office: Rethinking digital governance in the EU (Euractiv)… A Robust Governance for the AI Act: AI Office, AI Board, Scientific Panel, and National Authorities (academics on the EU Act)… Dutch data protection authority: scraping is almost always illegal… Public asked for their views on artificial intelligence (RTE)…

In the USA… There’s an AI Lobbying Frenzy in Washington. Big Tech Is Dominating (Time)… In DC, a new wave of AI lobbyists gains the upper hand (Politico)… The tech billionaires who helped ban TikTok want to write AI rules for Trump (Washington Post)… Biden administration rolls out international cybersecurity plan (Politico)… U.S. intelligence agencies’ embrace of generative AI is at once wary and urgent (PBS)… Department of Commerce Announces New Actions to Implement President Biden’s Executive Order on AI… Driving U.S. Innovation in Artificial Intelligence - roadmap from the Bipartisan Senate AI Working Group… Senators studied AI for a year. Critics call the result ‘pathetic.’ (Washington Post)… ‘I feel betrayed’: Schumer’s much-awaited AI road map gets panned by watchdogs (Fast Company)… Put the Public in the Driver’s Seat: Shadow Report to the US Senate AI Policy Roadmap… Why AI Policy Needs a Sociotechnical Perspective (Tech Policy Press)…

Everywhere else… How AI is impacting policy processes and outcomes in Africa (Brookings)… Who is in control of AI? (BBC)… Japan weighs regulating AI developers, following U.S. and EU (Nikkei Asia)… OECD updates AI Principles to stay abreast of rapid technological developments (OECD)… Politicians call for ban on ‘killer robots’ and the curbing of AI weapons (The Register)… The next wave of AI hype will be geopolitical. You’re paying (FT)…

And on a military theme… Military is the missing word in AI safety discussions (FT)… ‘Our Oppenheimer moment’ — In Ukraine, the robot wars have already begun (Politico)… What the viral “All Eyes on Rafah” meme means for the conflict (New Statesman)…

Deepfakes and democracy “I was the target of an AI nudes scam, and it could happen to you too – this is what everyone needs to know” (Stylist)… Tech Companies Promise to Try to Do Something About All the AI CSAM They’re Enabling (404 Media)… AI has created a new form of sexual abuse (Vox)… AI-Generated Child Sexual Abuse Material Is Not a ‘Victimless Crime’ (404 Media)… Call for safeguards to prevent unwanted ‘hauntings’ by AI chatbots of dead loved ones (University of Cambridge)… FKA Twigs uses AI to create deepfake of herself (BBC)… Schumacher’s family wins compensation for AI ‘interview’ (BBC)…

Could deepfakes influence the result of the world’s biggest election? (ITV News)… AI and deepfakes blur reality in India elections (BBC)… Indian Voters Are Being Bombarded With Millions of Deepfakes. Political Candidates Approve (Wired)… How China is using AI news anchors to deliver its propaganda (The Guardian)… Why China Is So Bad at Disinformation (Wired)… Moldova fights to free itself from Russia’s AI-powered disinformation machine (Politico)…

Politico’s Digital Bridge finds a more optimistic approach to AI and elections with CampaignLab… Don’t Hype the Disinformation Threat (Foreign Affairs)… Microsoft goes from bad boy to top cop in the age of AI and Anatomy of a scroll: Inside TikTok’s AI-powered algorithms as part of Politico’s Ballots and Bots series…

Generative AI and Democracy: Impacts and Interventions (Demos)… House of Lords debate… Tech pros call for rule change to allow media to refute AI-powered misinformation on election day (Public Technology)… Four steps to protect our elections from AI interference (Electoral Reform Society)… AI-Enabled Influence Operations: The Threat to the UK General Election (Centre for Emerging Technology and Security, Alan Turing Institute)… Disinformation: sources, spread and impact (POST)… ATTACK OF THE VOICE CLONES: How AI voice cloning tools threaten election integrity and democracy (Center for Countering Digital Hate)…

World of work Zombie Trainers and a New Era of Forced Labor (Newsweek)… AI and employment – the regulatory challenges laid bare (Diginomica)… How the computer games industry is embracing AI (BBC)… Pulitzer’s AI Spotlight Series will train 1,000 journalists on AI accountability reporting (Nieman Lab)… AI newsroom guidelines look very similar, says a researcher who studied them. He thinks this is bad news (Reuters Institute)… How news coverage, often uncritical, helps build up the AI hype (Reuters Institute)… How do people think #generativeAI will transform journalism and society as a whole? (Reuters Institute)… Newsrooms are experimenting with generative AI, warts and all (The Conversation)… AI aids England selection as squad for Pakistan series unveiled (BBC)… Will AI dream up the hit TV shows of the future? (BBC)… Writers and publishers in Singapore reject a government plan to train AI on their work (Rest of World)… Why mathematics is set to be revolutionized by AI (Nature)… Getty Images v Stability AI: the implications for UK copyright law and licensing (Pinsent Masons)… New York Daily News, Chicago Tribune, and others sue OpenAI and Microsoft (The Verge)… Sony Music slams tech giants for unauthorised use of stars’ songs (BBC)…

Big AI How the voices for ChatGPT were chosen (OpenAI)… Scarlett Johansson’s AI row has echoes of Silicon Valley’s bad old days (BBC)… OpenAI Just Gave Away the Entire Game (‘The Scarlett Johansson debacle is a microcosm of AI’s raw deal: It’s happening, and you can’t stop it’, writes the Atlantic)…

Introducing GPT-4o (OpenAI)… Today, we’re excited to introduce a new Gemini model: 1.5 Flash (DeepMind)… Wired’s take on these ‘inflection points’… Laughing, chatting, singing, GPT-4o is AI close to human, but watch out: it’s really not human (The Guardian)…

Our approach to data and AI (OpenAI)… Disrupting deceptive uses of AI by covert influence operations (OpenAI)… A landmark multi-year global partnership with News Corp (OpenAI)… A Devil’s Bargain With OpenAI: Publishers including The Atlantic are signing deals with the AI giant. Where does this lead? (The Atlantic)… unionized members of The Atlantic “are deeply troubled by the opaque agreement” (Axios)… OpenAI’s Long-Term AI Risk Team Has Disbanded (Wired)… Chaos and tension at OpenAI (Gary Marcus)… a thread from the outgoing head of alignment at OpenAI… Sam Altman’s take… AI firms mustn’t govern themselves, say ex-members of OpenAI’s board (The Economist)… OpenAI begins training next AI model as it battles safety concerns (FT)… Sam Altman was down and out at OpenAI. Now he’s taking on Google (Sunday Times)…

Providing further transparency on our responsible AI efforts (Microsoft)… Microsoft Build 2024: news and announcements from the developer conference (The Verge)… Microsoft’s new Windows 11 Recall is a privacy nightmare (BleepingComputer)… UK watchdog looking into Microsoft AI taking screenshots (BBC)… AI at Work Is Here. Now Comes the Hard Part (Microsoft) - thread from IFOW… Unlocking the UK’s AI Potential: Harnessing AI for Economic Growth (Microsoft, Public First)…

Hey Google: we need to learn more about your role in the war on Gaza (Access Now)… Introducing the Frontier Safety Framework (DeepMind)… Google scrambles to manually remove weird AI answers in search (The Verge)… CEO OF GOOGLE SAYS IT HAS NO SOLUTION FOR ITS AI PROVIDING WILDLY INCORRECT INFORMATION (Futurism)… Google Rolls Back A.I. Search Feature After Flubs and Flaws (Washington Post)… Google Researchers Say AI Now Leading Disinformation Vector (and Are Severely Undercounting the Problem) (404 Media)…

Friends From the Old Neighborhood Turn Rivals in Big Tech’s A.I. Race (New York Times with the latest Hassabis/Suleyman piece)… Jack Dorsey Leaves BlueSky Board and Calls X ‘Freedom Technology’ (Time)… Elon Musk’s Plan For AI News (Big Technology)… Slack under attack over sneaky AI training policy (TechCrunch)… How Dominic Cummings’ favorite AI firm captured the British government (Politico)… ‘I’m the new Oppenheimer!’: my soul-destroying day at Palantir’s first-ever AI warfare conference (The Guardian)… EU hits Meta with new probe over ‘addictive’ algorithms harming children (Politico)… For tech CEOs, the dystopia is the point (Brian Merchant)…

Everything else Science in the age of AI (Royal Society)… Using ChatGPT for analytics (Bennett Institute)… a new paper from Abeba Birhane and others, found that ‘as training data increased, so did the probability of misclassifying human images as offensive classes such as criminal’… No, Today’s AI Isn’t Sentient. Here’s How We Know (Time)… Meet the Woman Who Showed President Biden ChatGPT—and Helped Set the Course for AI (Wired)… 5 questions for UC Berkeley’s Ben Recht (Politico)… We’ll need universal basic income - AI ‘godfather’ (BBC)… The ugly truth behind ChatGPT: AI is guzzling resources at planet-eating rates (Guardian)…

Here’s how Network Rail is using that crazy AI camera tech (James O’Malley - see also his Test tube: The tube stations experimenting with AI for The House)… Britain’s got AI talent but that’s not enough (FT)… Governing in the Age of AI: A New Model to Transform the State (Tony Blair Institute, Faculty)… Artificial Intelligence (AI): The Future of Data (ONS)… We warn organisations must not ignore data protection risks as we conclude Snap ‘My AI’ chatbot investigation (ICO)…

Fully automated luxury mansplaining - How AI ruined Quora (Jonn Elledge)… What is AI? (Center for Countering Digital Hate)… Ethics of Artificial Intelligence: What it is and why we need it (John Tasioulas’s Elson lecture)… Surviving the AI Summer (Amy Bruckman)… Dynamics of Corporate Governance Beyond Ownership in AI (Common Wealth)… Can the AI future work for everyone? (FT)… The Problem With the Word ‘Alignment’ (AI Objectives Institute)… Building the Field of Democratic AI: Our Roadmap Launch and Webinar (The Collective Intelligence Project)… Today’s AI models are impressive. Teams of them will be formidable (The Economist)… ‘It might take over the world’ – schoolchildren issue warning over government AI (Public Technology)…

AI GOVERNANCE NEEDS SOCIOTECHNICAL EXPERTISE: Why the Humanities and Social Sciences Are Critical to Government Efforts (Data & Society)… Applying Sociotechnical Approaches to AI Governance in Practice (Center for Democracy & Technology - who’ve also published a report on Navigating Demographic Measurement for Fairness and Equity)… What does the public think about AI? US (Public First, Center for Data Innovation)… Releasing a new paper on openness and artificial intelligence (Columbia Institute of Global Politics and Mozilla)… Protesters Are Fighting to Stop AI, but They’re Split on How to Do It (Wired)… The harms of terminology: why we should reject so-called “frontier AI” (AI and Ethics, see also thread)… WTF is AI? (TechCrunch, though see also Emily Bender)… What Is AI’s Place in History? (Project Syndicate)… AI in public services: a conservative force? (Adam Lent)… wild, wild Westlaw (Deb Raji)…

And… really? REALLY?! Genius Group welcomes new Chief AI Officer, Alan Turing… Not a Genius move: Resurrecting war hero Alan Turing as your ‘chief AI officer’ (The Register)

DSIT up and take notice

There’s a lot from the UK government on AI Safety if you skipped the start of the ‘AI got ‘rithm’ section.

Sticking with AI… I put in a Freedom of Information request that revealed a promised inter-ministerial group on AI has apparently not yet been established… the NAO’s Chief Analyst delves into Use of AI in Government Report (Government Transformation)… IPA [Infrastructure and Projects Authority] Director on new AI project delivery framework (Government Transformation)… Our government shouldn’t use the public sector as a guinea pig for AI (The Hill)… Use of Entity Resolution in India: Shining a light on how new forms of automation can deny people access to welfare (Amnesty International)… and the ICO launched its fourth call for evidence on generative AI, this time on engineering individual rights into generative AI models…

Sticking with the ICO… Information Commissioner: Persistent sensitive information breaches failing people living with HIV (ICO)… while Watchdog under pressure to probe Met Police over use of facial recognition tech (the I) - coming as ‘I was misidentified as shoplifter by facial recognition tech’ (BBC), with more from Big Brother Watch… Learning from the mistakes of others – A retrospective review (ICO)… and PSNI facing a £750k fine following spreadsheet error that exposed the personal information of its entire workforce (ICO)…

Sticking with data breaches and cybercrime… Scottish national records data breached in NHS ransomware attack (Public Technology)… Birmingham Council-owned children’s services unit reprimanded over data breach of ‘criminal allegations’ (Public Technology)… MoD data breach: UK armed forces’ personal details accessed in hack (BBC)… Statement in response to reports of a cyber breach at the Ministry of Defence (ICO)… MoD hack: IT contractor concealed major hack for months (Computing)… Major government tech supplier SSCL faces probe after MoD data breach (Public Technology)… Data breach exposes details of 25,000 current and former BBC employees (Guardian)… though I wonder if this will go ahead, given the election - Exclusive: UK to propose mandatory reporting for ransomware attacks and licensing regime for all payments (The Record)…

To the Government Digital Service… GDS head Tom Read to step down (Public Technology)… GDS pledges user control, data minimisation, and ICO arbitration for new government digital ID service (Public Technology)… GDS creates CISO role to make services ‘radically easier and safer to use’ (Public Technology)… Using AI to generate web forms from PDFs (Tim Paul)… Work begins to reach new users via major ‘brand refresh’ for GOV.UK (Public Technology)… EXCL: New government digital strategy chief Gill sets data sharing and system reform as priorities (Public Technology)… that happened at Public Technology Live, where other highlights included ‘You have got to lead from the front’ – heads of major departments on their need for digital expertise (Public Technology)… and on digital government, there’s also the future of digital is already here – it’s just not evenly distributed (Matt Knight), Why fixing government tech is a nightmare (but not impossible) (Liz Lutgendorff) and This is the best thing the Tories did in office (James O’Malley)…

To statistics… Public confidence in ONS remains high as trust in civil service dips (CSW)… Public Confidence in Official Statistics (NatCen)… although Staff exodus hits UK’s statistics agency (FT)… and on evidence, Public site for government evaluations delayed due to resourcing issues (Civil Service World)…

Who could have possibly predicted this would happen… Worst-performing councils in England revealed — where does yours rank? (The Times)… LGA writes to Rt Hon Michael Gove MP on Oflog concerns (LGA)… thread from IfG’s Stuart Hoddinott…

Q: Can the UK be a ‘Science and Technology Superpower’?, asks RUSI (A: ‘the UK’s current science and technology policy is based on three fallacies that limit its ability to become a tech superpower’)… “Ground-breaking moment for science, innovation and technology” as UK’s most powerful supercomputer is officially online and debuts in global league (University of Bristol)…

And… Chancellor wants to build $1tn ‘British Microsoft’ (Computing)… Britain’s begging tech giants to list in London. Good luck with that. (Politico)… Independent review: Political violence and disruption (Walney Review includes bits on online safety)… Four projects shortlisted for implementation in 2024 Civil Service Data Challenge (Global Government Forum)… Controversial disinformation unit focused on ‘foreign states, election risks and deepfakes’ (Public Technology)… ‘Technical fault’ with online system allowed 67 couples to divorce early (Public Technology)… Local government bodies call for ‘centralised mechanism’ to support and fund AI (Public Technology)…

Improving Data in the Criminal Justice System (MoJ for Data in Government)… The Lady Chief Justice has created a new Transparency and Open Justice Board… Smarter regulation: delivering a regulatory environment for innovation, investment and growth (DBT)… Implementation of medical devices future regime (MHRA)… EXCL: DWP explores data science to inform Jobcentre ‘interventions’ for benefit claimants (PT)… DWP Pensions Dashboards programme adds £13m to core tech deal as report identifies ‘digital skills shortages’ (PT)…

And DSIT has published… explainers on the UK Product Security and Telecommunications Infrastructure (Product Security) regime (part of a ‘secure by design’ collection), smart devices and the Online Safety Act… a readout as AI Opportunity Forum holds penultimate meeting… details of webinars on the National Underground Asset Register… a statement on the adoption of UK Cyber Security Council standards, alongside Research on the cyber security of AI as Developers given new tools to boost cyber security in AI models as cyber security sector sees record growth - minister Saqib Bhatti also gave a speech on Improving UK cyber resilience: AI, software and skills… an invitation to join the Knowledge Assets Network for the public sector, alongside new resources… an announcement of the finalists for the Manchester Prize (for AI for public good)… a later deadline for Flexible AI Upskilling Fund pilot applications… and calls for evidence on Cyber security of AI, Code of Practice for Software Vendors, and the CyberFirst programme…

…while the pre-election period means that MP Stephen McPartland has published his DSIT-commissioned Review of Cyber Security and Economic Growth on his own website.

Parly-vous data?

Parliament was prorogued on 24 May and then dissolved on 30 May (by royal proclamation - there had been some disagreement as to exactly when). All business - bills, tabled questions etc - fell with prorogation, but select committees had some extra time to rush their reports out before dissolution. Those included the Education Select Committee, Screen time: impacts on education and wellbeing, summarised by the Guardian as MPs urge under-16s UK smartphone ban and statutory ban in schools (25 May) and the Science, Innovation and Technology Committee on the Governance of artificial intelligence (AI) (28 May). The SIT committee has helpfully published a summary of its activity during the parliament, and suggests some things its successors might want to look at.

Even before the election announcement, there was a lot going on.

Let’s start with the Commons… 2 May A PACAC report on lobbying noted the importance of better data, and how to handle messages on WhatsApp… 7 May Westminster Hall debate on cyber-security (Sir Mark Hendrick, Lab)… 8 May Women and Equalities Committee heard from campaigner and reality TV star Georgia Harrison plus OnlyFans CEO on tackling revenge porn… there was a ministerial statement on a Passport e-Gates Network Outage… 9 May MPs chose Jackie Doyle-Price as the new chair of PACAC (not that she had much time to enjoy it)… 14 May Westminster Hall debate on Smartphones and Social Media: Children (Miriam Cates, Con)… 15 May Ten Minute Rule motion on Child Sexual Abuse Material (Digital Devices)… 21 May The Women and Equalities committee noted that Government abandons pay transparency pilot… 22 May DSIT question time… Secretary of State Lucy Frazer was quizzed on AI and the creative industries as part of a ‘work of DCMS’ session (Q5 onwards gets straight to some of the big questions) - a few days before she spoke to the FT, UK working on rules for training of AI models with creative work… 23 May ministerial statement on the Seoul AI Summit… 24 May PACAC published their report on Transforming the UK’s Evidence Base (to which we contributed written and oral evidence)… 25 May a PAC report on whistleblowing included a bit on the need for better (or indeed, any) data…

As for the Lords… 2 May The government responded to the LLM report from the Digital and Communications Committee, which the committee was not delighted with, eliciting another response from DSIT… 7 May the European Affairs Committee took evidence on UK-EU data adequacy… 8 May a question on deepfakes and the general election (Maggie Jones, Lab)… and a statement on a defence data breach… 9 May Lord Holmes (Con) had a question on AI and intellectual property rights, while 10 May brought the third reading of his AI regulation bill… 14 May evidence from Ofcom as part of the future of news inquiry (Communications and Digital Committee - there was also a session on SLAPPs on 7 May)… evidence session on Data and Digital Trade… 15 May Question from Lord Davies (Lab) on Better, broader, safer: using health data for research and analysis (the Goldacre Review)… 21 May another evidence session on UK-EU data adequacy… 22 May The Joint Committee on National Security Strategy called on the Prime Minister to defend the UK’s democracy as snap election is called…

Separately, parliament published its Information and Digital Strategy for 2024-27.

We obviously don’t yet know what’s coming up or when, given the election - except that parliament will return on 9 July. With the number of MPs departing and a possible change of government, parliament - and all its select committees, whose constitution is determined by the election result - could look quite different.

In brief

What we’ve been up to

What everyone else has been up to

Events

Good reads

And finally:

 Previous

Need help?

Do you collect, use or share data?

We can help you build trust with your customers, clients or citizens

 Read more

Do you want data to be used in your community’s interests?

We can help you organise to ensure that data benefits your community

 Read more

Want to get involved?