Data Policy Digest

Gavin Freeguard

Gavin Freeguard

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

And Happy International Access to Information Day for yesterday! There is a lot of information to access in today’s edition, including a rundown of the various data and AI policy events at party conferences for those of you lucky enough to be going. I’ll be at Conservative, Green and Labour conferences if you want to say hi (assuming I recover from yesterday’s shocking news from Google).

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 Digest #1, Digest #2, Digest #3, Digest #4, Digest #5 and Digest #6, or sign up to the Digest and/or Connected by Data’s monthly newsletter via email.

Conference calls

Here are the data/AI/tech events happening at party conferences. You can skip over this section if beige buffets, warm white wine and cramped conference rooms are not your idea of fun.

Some conferences have happened already. The TUC Congress passed a motion on AI in the workplace which included a ‘legal right to consult trade unions on the introduction of new technologies is enshrined in law’.

A successful Lib Dem conference motion included pledges on ‘a clear, workable and well-resourced cross-sectoral regulatory framework for AI, that promotes innovation while creating certainty’, ‘establishing transparency and accountability for AI systems in the public sector’ and for the UK to play a ‘leading role in global AI regulation’. Big Brother Watch had an event asking if we need a digital bill of rights (they’re asking the same question at Conservative conference and I’d assume Labour, too).

Green Party conference is in Brighton from Friday 6 to Sunday 8 October. We have an event on the Saturday, ‘Protecting and empowering communities in an era of artificial intelligence’, with Natalie Bennett (former leader and now in the Lords), Andy Stirling (Professor of Science and Technology Policy at the Science Policy Research Unit, University of Sussex) and our very own Jeni. It takes place in Room 8 of the Conference, 17:15-18:30: to attend conference, including our session, book via the Greens’ website.

Conservative Party conference starts this weekend in Manchester. You can find full details on their website (one hopes the events will all be better designed than said website). Here’s a rundown of the events we’ve spotted - get in touch if we’ve missed anything.

SUNDAY ConHome and CityFibre discuss Digital Britain and connectivity… Airbus discuss the UK as science and technology superpower…

MONDAY Policy Exchange, Andreessen Horowitz and Coinbase ask how to make the UK’s digital sector the envy of the world… Centre for Social Justice and Virgin Money talk about digital exclusion… I chair an IfG/University of Surrey event on AI… Cliff42 wonder what AI unleashed could do for policing… Reform reimagine Whitehall… ConHome and Atos discuss digital skills… The Spectator asks if the Online Safety Bill is a threat to free speech… Onward and RELX have an event on the AI revolution and smart regulation… the Digital Poverty Alliance discuss digital inclusion… CPS and the City of London ask how the UK can become a tech superpower… Big Brother Watch ask if we need a digital bill of rights… the UK Tech Cluster Group wonder how to unlock the power of tech in the regions, at the Startup Coalition Tech Hub… the Startup Coalition then host drinks for young policy professionals with TBI…

TUESDAY DSIT Secretary Michelle Donelan appears on the main stage… during the day the Startup Coalition discusses AI Britain; the Digital Markets, Competition and Consumers Bill; the future of tech and foreign policy in the AI age; education in the age of AI; and host a tech leaders reception… techUK have events on AI resilience, the UK’s global role on AI and digital skills… Bright Blue and UK Stop Ad Funded Crime discuss cyber crime… Policy Exchange and Newton discuss how the use of data can influence successful implementation of Government policy… ConHome and Atos discuss digital transformation and public services… The Spectator and ABPI ask if AI and innovation could save the NHS… TBI discuss AI and national purpose… IfG and the British Academy consider the human side of AI… Digital Tories, Conservative Young Women and Conservative PCCs discuss online safety… ConHome, TrueLayer, NatWest and Open Banking have an event on Open Banking and the digital economy…

WEDNESDAY Policy Connect and the Social Care Institute for Excellence discuss social care and technology.

Labour Party conference gets going in Liverpool from Saturday 7 October. Slogan, apparently: Give Britain its Future Back.

Highlight, definitely: the Connected by Data event with the Fabian Society, ‘what Labour should (and shouldn’t) do on data and AI policy in the first 100 days’, with shadow DSIT ministers Chris Bryant and Matt Rodda, a stellar panel of civil society speakers and an audience packed with interesting data and AI perspectives. It’s 5.30pm on Monday 9 October at Tate Liverpool (outside the secure zone, so you don’t need a pass). The Fabians have a load of other events too, including one earlier on the Monday on the power of data and evidence-based public policy.

SUNDAY The PSC discuss public service delivery in the context of technological revolution… the Startup Coalition’s Tech Hub begins hosting a few days of events and space…

MONDAY Labour Together have a 7am(!) invite-only event on Labour’s vision for the digital future and a later one on innovation across the nation… I’ll be chairing an Institute for Government event on AI: governing the ungovernable? and later speaking at a private IfG roundtable on the human side of AI… Centre for Cities have an event on digital connectivity… the News Media Association discuss AI disinformation and the next election… the New Statesman have events with YouTube on harnessing tech for growth and with CityFibre on better connectivity… the Digital Poverty Alliance is discussing digital inclusion… the Campaign Lab have a campaigner’s guide to AI… Big Brother Watch ask if the UK needs a Digital Bill of Rights … Labour Digital discusses how tech can secure economic growth … Progressive Britain and the Solicitors Regulation Authority ask whether AI will revolutionise access to professional services… the Coalfields Regeneration trust discuss bringing tech and green jobs to the coalfields… the University of Liverpool discuss bridging the digital divide… the Startup Coalition, Taso Advisory, Labour Women in Tech and Labour Digital host the Labour Tech Reception…

TUESDAY starts with an event on whether technology in the workplace is a threat or an opportunity… Centre for Cities look at unlocking innovation across the UK… techUK and RELX discuss governing AI, one of several techUK events… Policy Exchange look at data and policy… Labour for the Long Term discuss Labour’s approach to AI… TBI ask if technology is a false hope or a new dawn… BT host a digital inclusion event… SME4Labour, TrueLayer, NatWest and Open Banking have an event on (surprise) Open Banking and the data economy… Microsoft asks if AI offers a new global role for Britain… BT discuss technology and the NHS… and Demos and AutogenAI host a discussion on generative AI and growth.

Well done on surviving that, and good luck for surviving conference season itself.

Data policy developments

Deeply DPDIB

We await the return of the Data Protection and Digital Information Bill at some point in the autumn - next up is Commons report stage, where MPs get to consider the amendments from committee.

Meanwhile… The government has published a statutory instrument on data protection, which means references in UK GDPR and the existing Data Protection Act will now be to the European Convention on Human Rights, rather than the Charter of Fundamental Rights of the European Union; that could have some impact on how some rights are interpreted and enforced… The outgoing Biometrics and Surveillance Camera Commissioner has reiterated that plans under DPDIB will leave an ‘oversight and guidance gap’ around biometrics… and in Europe, TikTok is apparently moving its data centres to avoid some regulatory challenges.

Bills, bills, bills

Online Safety Bill The Online Safety Bill passed gradually, then suddenly. After a year and half in parliament (and several years in consultation), the Bill sped through its final legislative stages and now awaits Royal Assent to officially become law.

Not before a whole other set of controversy and activity, of course - including a government climbdown that wasn’t, further warnings on investigatory powers and age verification, warnings from government of ‘humongous fines’ and No 10 hosting various celebrities (which I would put in quotation marks if it didn’t make me look/feel ancient and out of touch).

There are various reads on what happened and where we now are, including Politico focusing on some of the major players in the process, Wired saying the UK Is Poised to Force a Bad Law on the Internet, Carnegie UK publishing a guide to the Act (as it soon will be), and Michelle Donelan giving an interview to Hello! magazine. Ian Russell of the Molly Rose Foundation says the Bill will have failed if harm is not stopped; the New Statesman took a broader look at the (lack of) protection for children on YouTube.

But, as minister Lord Parkinson told the Lords: ‘I’m very conscious this is not the end of the road’. Attention will now turn to Ofcom and how the Act is enforced. The encryption debate will continue, not least through mooted updates to the Investigatory Powers Act and a Home Office campaign on the topic.

And in Europe, a new transparency database lists moderation decisions taken by major platforms under the Digital Services Act.

Digital Markets, Competition and Consumers Bill As with DPDIB, we await a date for Commons Report stage.

It’s been a busy few weeks nonetheless: the Competition and Markets Authority published some principles for markets around foundation models; the chair of the CMA gave a big speech about digital markets; the FT profiled the regulator; and across the pond there’s a big case brewing between the US government and Google.

AI Bill? Politico (paywall) quote a DSIT official ‘who has been in meetings about AI policy’ as saying ‘an AI Bill was not part of government policy “right now” and would not feature in the King’s Speech’. Though another person in those discussions said the debate was ongoing. Something to keep an (A) eye on.

AI got ‘rithm

It’s been another quiet few weeks on the AI Summit front.

Joking!

It’s appropriate that in its blurb about the Summit, DSIT describes us as being at a ‘crossroads’, since the arrangements continue to resemble a shoddily constructed soap opera at times.

But there has been some progress. There’s now a page collecting all the news about the Summit. We have some further information on scope - the Summit will focus on misuses or loss of control risks around ‘frontier AI’, defined as ‘highly capable general-purpose AI models that can perform a wide variety of tasks and match or exceed the capabilities present in today’s most advanced models’, and narrow AI with dangerous capabilities such as bioengineering. The Ada Lovelace Institute called for a much broader definition of AI safety in their handy guide to a successful summit.

And we have details on ‘engagement with stakeholders’ before the Summit. There will be four (apparently invitation-only) pre-summit events hosted by academia and industry groups: the Turing on UK strengths and opportunities for international collaboration (11 October); the British Academy on possibilities of AI for the public good (12 October); techUK on opportunities, risks and solutions in the tech sector (17 October); and the Royal Society on horizon scanning AI safety risks across scientific disciplines (25 October). There will be two online Q&As: with Matt Clifford, one of the Prime Minister’s representatives to the Summit, on Twitter/X (2 October - he’s taking questions already); and with DSIT Secretary of State Michelle Donelan on LinkedIn (18 October). You may well wonder to what extent that, and being able to watch a livestream of keynote speeches from the Summit on 1 November, constitutes stakeholder engagement.

There will be an AI Summit Fringe, being put together by various industry, academic and civil society groups from 30 October to 3 November. Further details to follow, but it looks like that will have much wider representation of groups and communities affected by AI and wider coverage of AI-related topics than the Summit itself.

China apparently have been invited to the Summit proper despite political pressure and rumours they might only be invited to part of it. Several EU nations have apparently been left out, and I don’t think I’ve seen any reference to invitees from the Global South. Politico noted international scepticism earlier this month (their paywalled Morning Tech newsletter has even choicer quotes - one European Commission official: “No one is really sure what the U.K. AI summit is about… We’ll likely send someone, but how senior that person is depends on how much the [Commission] president wants to put into this’). Quote of the week, from a UK briefing in Brussels: MEP and EU AI Act champion, Dragoș Tudorache, asking “Do I need to actually go and say I can build a bomb, in order to be invited?”

The Future of Life Institute called for more civil society representation (among other things) while The Times reports fears that big tech companies will take over AI safety conference, including from industry. (Over in the US, there are also concerns that key voices - notably workers - are missing from the debate, with Bloomberg and the BBC among those covering various meetings.)

Both Politico and the Telegraph (‘this is his climate change’, unlike actual climate change) say Prime Minister Rishi Sunak wants AI to seal his legacy. He was apparently angered by an FT journalist asking why he thought a “midsize country” might be in a position to write the global rules for artificial intelligence. There’s a lot of concern about the PM’s position on AI: he’s made no secret of the Silicon Valley influence, but Politico has an extensive read on the ‘Silicon Valley doomers’ shaping his view, which has led to the narrow, existential risk-based definition of AI Safety for the Summit. (The Guardian also touches on the debate.) Politico thinks the tech bros have already won and notes the irony of Silicon Valley execs, shunned over social media failures, ‘now being embraced as soothsayers who are the only ones with the technical knowledge to quell the long-term concerns associated with advanced AI’.

Meta’s Yann LeCun diagnosed Sunak with ‘Existential Fatalistic Risk from AI Delusion disease (EFRAID)’. Be efraid, be very efraid: there’s been a notable emphasis on AI and bioweapons from the UK government this week, not just from Sunak, but from DSIT secretary, Michelle Donelan. She also addressed the CogX conference earlier this month (not that there were many people there to hear it) and attended a generative AI hackathon in Downing Street. The Frontier AI Taskforce (formerly the Foundation Model Taskforce) published its first progress report alongside a video from taskforce chair, Ian Hogarth (profiled by The Times recently), and a press release. There’s also a new advisory service to help businesses launch AI and digital innovations, and some case studies on AI assurance.

Back on the international stage, deputy PM Oliver Dowden spoke at the UN about AI being ‘the biggest transformation the world has known’. The UN is itself considering how to shape AI, perhaps through a new agency, while the G20 will (among other things) ‘pursue a pro-innovation regulatory/governance approach that maximizes the benefits and takes into account the risks associated with the use of AI’. (Japan, which is leading the G7 process on AI, is considering its own domestic rules on business transparency, too.)

European Commission President Ursula von der Leyen got existential in her state of the union address, too. She also wants Europe (and partners) to ‘lead the way on a new global framework for AI’ built on guardrails (including the AI Act), governance (including a possible equivalent to the IPCC in AI risks and benefits) and innovation (including opening up ‘our high-performance computers to AI start-ups to train their models’). IAPP have a timely guide to the EU’s AI Act as Avaaz launch a campaign to put human rights at the forefront of it. Raymond Sun has put together a global AI regulation tracker, as the BBC asks whether it’s even possible to regulate AI.

Elsewhere… debates continue about AI, copyright and the creative industries, with Creative Commons publishing an Open Letter from Artists Using Generative AI, George R R Martin and John Grisham among the creators launching a lawsuit, and Music Business Worldwide taking a closer look at one of the existing cases involving Sarah Silverman. Getty Images have made an AI generator only trained on its licensed images.

On the business side, Amazon have invested in Anthropic (who published a ‘responsible scaling policy’, not dissimilar to the ‘responsible capability scaling in Donelan’s CogX speech), while Wired took a look at What OpenAI Really Wants (Nirit Weiss-Blatt shared what OpenAI’s chief scientist said at an event a few months ago). Their ChatGPT can now access up-to-date information. Norway’s oil fund is sending a message to companies on AI, according to the FT (full guidance also available).

And… Nature looked at spotting ChatGPT use in scientific papers, and UCL at AI providers, essay mills and contract cheating law… Ethan Zuckerman asked What happens when AI trains itself?… the FT explained how generative AI works… TIME published a list of 100 influential people in the field… the Bishop of Leeds shared his thought for the day on AI… techUK launched a new report on AI and work… and MIT marked six months on from the open letter calling to pause AI development.

DSIT up and take notice

Activity, appointments and… abolition galore at various arm’s length bodies. The advisory board to the Centre for Data Ethics and Innovation was quietly disbanded, joining the AI Council - CDEI has now published some reflections from departing board members. The Advanced Research and Invention Agency, meanwhile, has announced its programme directors (and done a short Twitter video with them). Dr Dave Smith is the new national technology adviser, while Dean Creamer is the new chief exec of Building Digital UK.

The big news is the UK agreeing a data bridge with the US (a data bridge being ‘granted by the UK to individual countries, sectors or international organisations which provide adequate standards of protection for personal data’ without the need for contractual safeguards). The ICO has published an opinion (tl;dr: reasonable to think data protection is adequate, but the government should monitor data on criminal offences, automated decision making and the right to be forgotten). Computer Weekly has more. The UK also signed a memorandum of understanding with Israel.

And… the UK agreed a bespoke deal to rejoin the EU’s Horizon and Copernicus research programmes (to much relief)… the UK’s new supercomputer will be built in Bristol… the department joined a cross-Whitehall scheme to drive science and tech expertise among civil servants… opened a consultation on ‘Open Communications: a Smart Data scheme for the UK telecoms market’… published details of successful ‘smart infrastructure’ pilots… and there’s a new competition for accelerating the use of quantum technologies in the public sector…

Outside of DSIT, chancellor Jeremy Hunt spoke to Politico about his trip to the US West Coast to talk tech, one of his big selling points being the UK’s post-Brexit ‘regulatory autonomy’ and the government’s desire to have ‘the smartest, pro-innovation regulatory regime incredibly quick and fast’. (Royal Assent for the Online Safety Act will come four years after the online harms white paper was published, it’s over a year since the first version of DPDIB got lost somewhere in the Truss takeover, and it’s three months since the AI Regulation white paper consultation closed and we have no idea whether the government’s obsession with big AI’s talking points means a change of course, in case you were wondering.)

And the IfG’s latest batch of Ministers Reflect interviews include Chloe Smith - complete with details of her maternity cover for Michelle Donelan as DSIT secretary of state - and former tech minister Matt Warman.

Parly-vous data?

The Public Accounts Committee published a report on digital within government: ‘Whitehall staffing cuts add to digital skills shortages and risk increased costs’. Otherwise not a huge amount in the brief sitting between summer and conference recesses, but remember there are active select committee inquiries into AI governance, LLMs, AI in weapon systems, the future of transport data and transforming the UK’s evidence base.

The next DSIT question time in the Commons is scheduled for 1 November - also the first day of the AI Summit. Though that assumes parliament doesn’t wrap up before that, ahead of the State Opening of the next session on 7 November.

Labour movement

Our last edition dropped part way through Labour’s frontbench reshuffle. We now know the full line-up, which is:

  • Peter Kyle Shadow Secretary of State for Science, Innovation and Technology
  • Chi Onwurah Shadow Minister for Science, Research & Innovation
  • Chris Bryant Shadow Minister for Creative Industries and Digital
  • Alex Davies-Jones Shadow Minister for Tech and Digital Economy
  • Matt Rodda Shadow Minister for AI and Intellectual Property
  • Lord (Steve) Bassam Shadow Spokesperson
  • Lord (Wilf) Stevenson Shadow Spokesperson.

We think Chris will have responsibility for the Data Protection Bill, and that Jonathan Ashworth - reshuffled into the shadow Cabinet Office team as shadow paymaster general - will be thinking about the role of data in government. And if you want to know more about Peter Kyle, he’s done a very moving interview with The Times - this gives you a flavour.

Labour’s recent prime ministers have been talking tech this month. The FT interviewed Tony Blair, who says that rather than tax and spend, ‘the radical agenda today is all about understanding, mastering, harnessing the technological revolution — everything else is secondary to that’. With the tech revolution coming anyway, politicians need to regulate it, as they did with health and safety rules during the industrial revolution. Gordon Brown, meanwhile, has been talking about AI boosting the economy (though I’m not entirely sure how).

In other news… The Times had some more about Sue Gray’s possible role in ‘perform[ing] open heart surgery on the British state’ and Starmer’s conception of politics as ‘institutional reform’ (he ended his tenure as Director of Public Prosecutions ‘recommending the CPS rethink its internal filing system to reduce its reliance on paper’)… the National Policy Forum ‘agree[s] with the OECD that collective bargaining can help companies and workers respond to demographic and technological change and adapt to the new world of work’… pre-conference pieces are appearing on the influence of Labour (and Tory) thinktanks on the party’s thinking… and there are some rumours that we could be in for a May election (a reminder - it has to take place before the end of January 2025).

In brief

What we’ve been up to

What everyone else has been up to

Events

If all those conference events haven’t broken your calendars…

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?