Data Policy Digest
Hello, and welcome to our twentieth Data Policy Digest, bringing you all the latest data and AI policy developments.
Going through the motions, too cautious, little spark, relief that it will soon be over, and little mention of data and tech… enough about England and Scotland at the Euros, the general election is a week today. And civil society tried to inject some data and AI discussion into the campaign with two events this week - more below.
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
What everyone else has been up to
Never mind the ballots
If you’ve not already bookmarked it/made it your homepage/pledged to vote for it on 4 July, we’ve set up a Data and AI Civil Society Network Election Hub. We’ll be updating with resources throughout the campaign - and you can add to it, too! (And if you’d like to find out more about the network, visit our dedicated website.)
Manifesto destiny
On Monday, the Democratic Unionist Party launched their manifesto. There’s a short section on AI, about ‘harness[ing] the potential of Artificial Intelligence and smart data to build stronger communities and drive economic growth’ while recognising ‘significant risk. We will oppose plans to allow AI developers to use existing music, literature and works of art without permission or payment and disregarding copyright laws. DUP MPs will also seek more robust, but proportionate, regulation of data and emerging technologies in order to crack down on abuse, harmful practices such deepfakes and efforts to undermine democracy through misinformation and disinformation.’… there’s a pledge to campaign for a ‘national database of public land’… another to establish ‘clusters of universities, business and clinicians throughout the UK to harness and accelerate the use of new medical technology and diagnostics’ and ‘support accessible and portable health information throughout a patient’s life’… and close to a whole page (p43) on Online Harm, where the promise to ‘campaign to end the ‘Wild West’ of social media and to protect children and all internet users from harm, abuse and misinformation online’ includes a ‘rapid review of emerging risks from AI’ and supporting progress on ‘stronger rights for bereaved parents in accessing their child’s social media profiles and data’.
I can’t actually find the SDLP manifesto, launched on Wednesday, online - though the BBC summary doesn’t have anything obviously data-relevant. I couldn’t find anything in the Sinn Féin manifesto either.
While we’re on manifestos… Defend Digital Me has scrutinised party plans on digital and education… the Digital Poverty Alliance has looked at the Conservative (‘All Quiet on the Digital Poverty Front’), Labour (‘Taking the Poverty out of Digital Poverty’) and other manifestos… and there’s also Assessing the manifestos against our calls for change (Full Fact)… and 2024 UK General Election: 4 ways the next government can make the internet safer (Center for Countering Digital Hate).
Labour movement
How might a Labour government reorganise the machinery around data and digital? Will Perrin (ex-No 10 under a previous Labour government) unpacks Peter Kyle’s London tech week speech (Will Perrin). He also refers to a Wes Streeting appearance at a Medical Journalists’ Association event: Digital Health has most on data, Wes Streeting slams ‘glacial’ pace of FDP adoption, though the MJA also links to other coverage. Global Counsel also has a podcast, on How will a Labour government approach tech policy?
If you’re interested in workers’ rights on data and AI… Labour has previously committed to a Single Enforcement Body, which will now be called the Fair Work Agency: Labour watchdog will have ‘real teeth’ to prosecute rogue employers, says Angela Rayner, according to The Observer. Let’s see if that includes digital rights.
And Politico also highlighted a Labour press release about R&D plans, which includes AI as one of six priority areas (alongside aerospace, automotives, defence, energy and life sciences). Not sure that bit has been written up anywhere not paywalled, but Yahoo News carry the broader Press Association report.
News Tory
Is there any point, really?
Everything else
Tuesday was civil society data and AI election event extravaganza day!
It started at 4.30pm with our Question Time-style debate, Who should Labour listen to on AI, tech and public services?, with our own Jeni Tennison, the Tony Blair Institute’s Jeegar Kakkad, Nesta’s James Plunkett and Public Law Project’s Caroline Selman, chaired by the FT’s John Thornhill. You can read a summary or watch it back in full.
Then, in the evening - and infinitely preferable to any football that may have been on, or so I understand - came a hustings on ‘digital rights in the age of AI’, organised by ORG with us, Big Brother Watch, Defend Digital Me, the Open Data Institute, Worker Info Exchange and Institute for the Future of Work. Author and broadcaster Timandra Harkness chaired Viscount Camrose for the Tories, Lord Clement-Jones for the Lib Dems, and Cllr Jack Lenox for the Greens, Labour conspicuous by their absence. You can also watch that back in full.
Elsewhere… AI and IP apparently came up at the Creative UK hustings (h/t Politico), with Labour’s Chris Bryant pointing to Labour’s Plan for the Arts, Culture and Creative Industries (try page 6 for the headline and page 17 for more)… the Online Safety Act and scams came up at the MoneySavingExpert hustings (again, h/t Politico)… the Royal Statistical Society has asked parties to commit to basic stats training for ministers (Civil Service World)…
Channel 4 asks ‘Can AI steal your vote?’… Full Fact have been using AI to support their debate factchecking (Politico)… though Politicians’ claims rather than AI fakes have kept UK election fact-checkers busy (Press Gazette)… and Politico’s Digital Bridge also struggles to find deepfake impact (while its Influence newsletter looks at the effect video doorbells are having).
Data policy developments
AI got ‘rithm
How AI is changing warfare (The Economist)… How to Fix “AI’s Original Sin” (Tim O’Reilly)… AI IS ALREADY WREAKING HAVOC ON GLOBAL POWER SYSTEMS (Bloomberg)… The Hidden Environmental Impact of AI (Jacobin)… Singapore’s ‘Fomo’ amid the rise of the AI data centre (FT)… Algorithms should not control what people see, UN chief says, launching Global Principles for Information Integrity (UN)…
We’re sharing an update [read: delay] on the advanced Voice Mode we demoed during our Spring Update (OpenAI)… Anthropic has a fast new AI model — and a clever new way to interact with chatbots (The Verge)… Apple Won’t Roll Out AI Tech In EU Market Over Regulatory Concerns (Bloomberg)… Apple, Meta not in talks currently for AI partnership, Bloomberg News reports (Reuters)… EU accuses Apple of breaching landmark digital law (Politico)…
Generative AI Can’t Cite Its Sources: How will OpenAI keep its promise to media companies? (The Atlantic)… AI Tools Make It Easy to Clone Someone’s Voice Without Consent (Proof)… World’s biggest music labels sue over AI copyright (BBC)… YouTube in talks with record labels over AI music deal (FT)…
AI Procurement Checklists: Revisiting Implementation in the Age of AI Governance (Computers and Society)… Citizens’ data afterlives: Practices of dataset inclusion in machine learning for public welfare (AI and Society)… Most small charities are using AI, report finds (Civil Society)…
And… ChatGPT is bullshit (Ethics and Information Technology). And if that language isn’t strong enough for you…
Everything else
DSIT up and take notice In Broken Britain, even the statistics don’t work (FT)… ICO police cloud guidance released under FOI (Computer Weekly)… DWP algorithm wrongly flags 200,000 people for possible fraud and error (Guardian)… and AI is one of the issues covered here by Sir Patrick Vallance: “If you’re a regulator, the incentive to take risks is virtually zero” (Andrew Bennett, The Form Playbook).
Bills, bills, bills Enjoy the parliamentary peace while it lasts… in the meantime, an FOI request gives a sense of the internal ICO reaction to the fall of the Data Protection and Digital Information Bill.
Parly-vous data? Parliament will meet after the election on 9 July, when MPs will be sworn in and the speaker elected. The state opening and King’s Speech - which will outline the government’s legislative agenda - will take place on 17 July.
In brief
- Russian hackers behind NHS attack are part of Kremlin-protected cyber army (the I)…London hospitals hackers publish stolen blood test data (BBC)… All 150,000 residents of Dumfries and Galloway warned to assume data loss in NHS cyberattack (Public Technology)… Three things we must do now to prevent patient harm from digital tech (Digital Health)
- ID Verification Service for TikTok, Uber, X Exposed Driver Licenses (404 Media)
- Biden admin can’t be sued for pressuring social media companies to remove misinformation, Supreme Court rules (Politico)
- Australia’s eSafety commissioner waters down child abuse detection rules in online safety standards (The Guardian)
What we’ve been up to
- A reminder of our new(ish) project on how the public sector should give the public a say on data and AI. As part of that we’re starting a peer network to bring together practitioners from across the public sector - and other sectors. We want to make sure it’s actually helpful, so we have a survey. This is the survey if you work in the public sector - this is the one if you’re somewhere else. Thank you!
- See above for more on our events on Who should Labour listen to on AI, tech and public services? and the hustings on ‘digital rights in the age of AI’
- We published weeknotes by Jeni and Emily
- On 9 July we have a workshop, Scoping training needs on AI for trade unionists in Wales, with Wales TUC
- On 18 July, Tim will be facilitating a workshop on ‘A permanent Global Citizens’ Assembly: Adding humankind’s voice to world politics - AI’, as part of a conference on the establishment of a permanent Global Citizens’ Assembly (GCA), organized by the University of Oxford and Iswe Foundation
- On 10 September, we have a Connected Conversation on ‘How can we ensure meaningful community and public participation in the governance and implementation of the EU’s AI Act?’
What everyone else has been up to
- The Digital Trade & Data Governance Hub at the George Washington University have published a new report as part of their Data Governance Mapping Project
- The ODI have been busy, with a new report on What makes participatory data initiatives successful?, a series of reports on ‘Understanding data governance in AI’, and posts on Enforcing people’s rights in the data supply chain, AI data transparency: understanding the needs and current state of play, How can we empower data communities in the era of generative AI?, Transforming AI data governance with Croissant: a new standard for ML metadata and Why the UN Global Digital Compact needs to get serious about participatory data.
- IFOW have a new briefing paper, What impact does exposure to workplace technologies have on the quality of people’s jobs? (On a similar note: BCG also have a report out, AI at Work: What People Are Saying)
- People vs Big Tech have launched a newsletter - the first edition was on The People vs Misogyny
- Peter Wells and IF published A toolkit to help create more trustworthy advertising experiences
- 360Giving launched UKGrantmaking 2024
- ORG are advertising for a couple of positions: technical investigator on election apps and a consultant on data protection
Events
- 1-7 July: London Data Week…
- 2 July: …which will include me presenting episode 4 of The Data Game as part of the ODI’s Data Time TV
- 9-11 July: Data for Policy 2024 conference
- 11 July: Data-centric AI webinar #2: the data scraping challenge (ODI)
- 15 July: The first year: what the new Government should do on data and AI (ODI)
- 16 July: Now the Work Begins: Lessons in a New Kind of Democracy with Audrey Tang (Demos)
- 16 July: Civic AI Unconference
Good reads
- A Search for the Heart: Digital collections, machine learning and legacies of care (LIbrary of Congress)
- How the Square Root of 2 Became a Number (Quanta)
- THE ENCYCLOPEDIA PROJECT, OR HOW TO KNOW IN THE AGE OF AI (Public Books)
- Are there any humans left on the internet? (Prospect)
- Tony Blair - Life of a PM, The Deep State, Lee Kuan Yew, & AI’s 1914 Moment (podcast and transcript - Dwarkesh Patel)
And finally: