An invitation to the world’s leaders to attend an AI summit in the UK was a big play by the prime minister – but Rishi Sunak can justifiably point to some successes, writes Matthew Gill
The prime minister would have been pleased with last week’s artificial intelligence (AI) safety summit. He secured some high profile political figures – including US vice president Kamala Harris and EU commission president Ursula von der Leyen – as well as tech leaders like Demis Hassabis, Sam Altman and Elon Musk. Sunak also appears to have ridden out the controversy that China’s attendance could have caused, with Musk describing the Chinese involvement as “essential.” 17 https://twitter.com/RishiSunak/status/1720187297558065441
Nonetheless, to a non-Silicon Valley native who did not believe that tech companies rule the world, it would all have seemed quite strange. The spectacle of a prime minister interviewing a tech entrepreneur on that entrepreneur’s own social media channel – soliciting the entrepreneur’s predictions for the future, letting his blithe assertions about work becoming optional and AI replacing human friendship go unchallenged, and seeking his approval for the decision to invite China to the summit – might seem, to the uninitiated, to devalue the office of PM. But while it is fair to ask whether No.10 really needed to yield so much of the framing to Musk just to secure the interview, Sunak can point to a summit that achieved aims, albeit narrow ones.
The optimism of the summit – and optimistic coverage – was not a foregone conclusion
A successful summit was by no means a foregone conclusion. It was arranged quickly, but key delegates were not confirmed until very late in the day; the USA’s assertive positioning in the run-up could have made the UK look lightweight; and the conversation with Musk was oddly deferential. The UK’s aspirations to global leadership were humoured as much as embraced – and there was always the sense that they may not have been. It is not clear that we can expect to be humoured again.
But the prime minister can feel satisfied with how the Bletchley summit played out. Having said beforehand that “the overarching risk is a loss of trust in and trustworthiness of this technology which would permanently deny us and future generations its transformative positive benefits”, 18 26 October 2023, Gov.uk, 26 October 2023, www.gov.uk/government/speeches/prime-ministers-speech-on-ai-26-october-2023 prime time news coverage dominated by driverless cars and potential cures for cancer powered by AI – rather than by tales of large scale human obsolescence – would have felt like a clear win.
Sunak is in his element talking about technology, innovation, and particularly enterprise finance, but capturing the public’s imagination in these areas without seeming out of touch is a fine line to tread. There is much for the government to do domestically to ensure that AI can be adopted ethically and effectively – including in a way that respects data privacy, avoids bias and minimises misinformation – but for now Sunak will be happy with a relatively optimistic public narrative.
Concrete announcements on AI were not guaranteed before the summit
The PM can also point to some concrete announcements 24 2 November 2023,Gov.uk, 2 November 2023, www.gov.uk/government/speeches/prime-ministers-speech-at-the-ai-safety-summit-2-november-2023 mixed in with the warm words of the Bletchley Declaration 25 Department for Science, Innovation & Technology, Foreign, Commonwealth & Development Office and Prime Minister's Office, 10 Downing Street, The Bletchley Declaration by Countries Attending the AI Safety Summit, 1-2 November 2023, Gov.uk, 1 November 2023, www.gov.uk/government/publications/ai-safety-summit-2023-the-bletchley-declaration/the-bletchley-declaration-by-countries-attending-the-ai-safety-sum… that was signed by summit participants:
- A multilateral agreement for governments to test AI models – and for developers to submit their models to testing;
- A permanent AI Safety Institute in the UK, which will work collaboratively with – amongst others – the safety institute announced earlier in the week by the USA (notwithstanding the debate as to whether the UK was upstaged by the USA’s announcement)
America is a friend, it’s not trying to ‘upstage’ the UK’, Financial Times, 3 November 2023, www.ft.com/content/5d0561d8-e42b-4c63-a0b6-e926b3b9c877
- An international advisory panel to advise on frontier AI risk, modelled on the Intergovernmental Panel on Climate Change;
- Agreement that the advisory panel, will support production of a “State of the Science” report, chaired by Yoshua Bengio with a UK secretariat, ahead of the next international summit;
Statement by the Chair, 2 November 2023, Gov.uk, 2 November 2023, www.gov.uk/government/publications/ai-safety-summit-2023-chairs-statement-state-of-the-science-2-november/state-of-the-science-report-to-understand-c…
- Announcement that a series of future AI summits will take place, hosted by other countries.
Two weeks ago, this package of announcements – while relatively limited in focus – would have looked like the upper end of plausible outcomes from the summit, so they should be chalked up as a success.
The high point of UK influence in AI may already have passed
The rest of the world is now picking up the mantle – swiftly – from the UK. South Korea will host another summit in the first half of 2024; France a year from now. Current voluntary arrangements for research into emerging AI models could quickly morph into regulatory requirements in the USA and Europe. Despite being one of the leading centres for AI, the UK’s chances of leading the development of a regime that will include the USA – where most of the leading AI firms are based – and China – which has greater economic power – seem small. Its aim must be to retain an influential and constructive seat at the table, rather than to dominate.
But even geopolitics itself is perhaps a sideshow. The catalysts to AI’s development are technical and commercial, not political. The AI summit has successfully brokered a global governmental conversation about risks and rules, for which Sunak – and the UK – deserve credit. But governments are still playing catch-up.
In this context, the UK’s government and politicians will need to play their hands realistically if they are to continue to make a constructive contribution to the developing global conversation and to ensure that the key questions of the rights and responsibilities that should be associated with AI are properly and publicly addressed. On the one hand, providing a flexible, imaginative alternative to a rules-bound EU in this regard is precisely the kind of thing that a post-Brexit UK would want to show that it can do; but on the other hand the practical influence of that alternative will be weaker now that the UK sits outside the larger trading bloc.
For Sunak, though, the summit can be considered a brief success – albeit not one that is likely to be much noticed by the electorate. The effort and energy that went into organising the two-day gathering suggest a PM with one eye on his legacy – and perhaps another on his future. Sunak will want to remain as prime minister for many more years, but if he were to lose the next election then some have suggested he might see a role in Silicon Valley, following in former deputy PM Nick Clegg’s footsteps, as his next move. 28 the prime minister was more starry-eyed than a SpaceX telescope’, Guardian, 3 November 2023, www.theguardian.com/commentisfree/2023/nov/03/rishi-sunak-elon-musk-prime-minister-spacex-silicon-valley That may or may not be in his mind, but the very idea that it might constitute a suitable next step for a former prime minister of the UK perhaps reveals a shift in the standing of the UK and its political establishment. Things would have looked quite different only a relatively few years ago.