Sunak’s AI summit to prove biggest test of UK’s relationship with Beijing
James Titcomb
6 min read
Rishi Sunak has expended significant political capital in trying to ensure his AI summit is a lasting success - NEIL HALL/EPA-EFE/Shutterstock
The bucolic setting of Bletchley Park, the home of British codebreaking operations during the Second World War, is a fitting backdrop to Rishi Sunak’s ambitions to address a new threat: AI.
In two months, the Buckinghamshire country estate will host the Prime Minister’s AI Safety Summit, a first of its kind international effort to ensure that the risks of rapidly improving artificial intelligence are addressed.
Sam Altman, the chief executive of ChatGPT owner OpenAI, has said that the technology needs a global regulator similar to the International Atomic Energy Agency (IAEA).
Sunak has expressed sympathy for the idea and would like to see it set up in London.
However, there is an elephant in the room. Ministers are yet to confirm whether China will be invited to the summit.
A regulator that did not have Chinese endorsement could not realistically claim to be global. Yet China’s involvement is a delicate issue.
Inviting representatives from Beijing would risk angering allies such as Japan and the EU, and possibly even Britain’s own security services. China’s growing technological sophistication is seen as a very real threat to Western values.
Yet not to include China would risk undermining Sunak’s ambition to establish Britain as the centre of AI regulation.
China ranks second in the world for artificial intelligence funding, behind the US, and leads in academic publications, according to Stanford University’s Institute for Human-Centered Artificial Intelligence.
The Communist Party has said it wants the country to be the world’s leading AI hub by 2030 and the country is home to AI giants such as SenseTime, Baidu and Tencent.
Sunak must figure out how to reach international consensus on AI against the backdrop of an escalating technology war between Beijing and the West.
A recent British defence review classed China as an “epoch-defining challenge”. GCHQ has warned of growing threats from Chinese cyber attacks, which are likely to be amplified by AI that can convincingly trick humans into handing over security details.
Chinese officials have insisted that AI in the country acts effectively as an arm of the state. The country’s powerful cyberspace administration has ordered that Chinese chatbots “adhere to the core values of socialism” and not encourage attempts to overthrow power.
In response, the US has barred the semiconductor giant Nvidia from supplying its most advanced AI chips to China and is expected to issue further crackdowns on new versions the company has released to serve the Chinese market.
The FBI has accused China of stealing American AI technology to “enable foreign influence campaigns”.
A US National Security Council spokesman told Politico last week that the US would be “fine with China attending the summit”.
However, Japan, another key security ally and a key player in the development of AI, has reportedly opposed the country being invited, while European Union officials are also sceptical.
A report from MPs on the Science and Technology Committee released on Thursday calls for invitations to the summit to “be extended to as wide a range of countries as possible”.
However, it also called for a separate forum of democratic countries that would “ensure mutual protection against those actors – state and otherwise – who are enemies of these values and would use AI to achieve their ends”.
Greg Clark, the committee’s chairman, said that the West should be able to work with China on certain AI safety goals, drawing comparisons with international efforts to prevent toxins in the development of medicines.
“The technology is going to be global,” Mr Clark said. “And there is some thinking to be done about AI safety across all countries and we should try to explore whether it is possible to have an agreement on this. If this is to be the first global AI summit, to have as many voices as possible would be beneficial.”
However, he said that on matters of national security, “It would appear to be wrong to have those discussions in the company of people who didn’t share your values and worldview.”
Foreign Affairs select committee head Alicia Kearns expects China to be invited but for ‘limits to be placed on their participation’ - Heathcliff O'Malley
Alicia Kearns, the Conservative head of the Foreign Affairs Committee, said: “I would expect China to be invited [to the summit], but limits to be placed on their participation given they are not a trusted partner.
“China is an AI competitor to the UK, where possible we should seek to engage with them to bring them on board with the need to recognise global regulations on AI and manage the worst excesses of the ways in which they may seek to use AI to undermine other states and actors.”
Kearns said that China’s “future behaviour will determine the extent to which cooperation can continue”.
James Cleverly, the Foreign Secretary who is this week visiting China in the first such trip for five years, declined to say whether Beijing officials would be invited to the Bletchley Park summit but pointed out that he had invited Chinese experts to speak at the UN Security Council.
Sunak himself has been non-committal on the issue. Last month he told MPs on the Liaison Committee that “AI does not respect national borders, and I think we will all benefit from hearing and talking to each other,” but added that the first step was discussing the matter with “like minded countries”.
Cleverly told the Financial Times that the guest list for the summit, scheduled to be held on November 1 and 2, is still being finalised.
Foreign Secretary James Cleverly has not yet said whether Beijing officials would be invited to the Bletchley Park summit - Rory Arnold/No10 Downing Street
One official familiar with the summit’s planning says that China is almost certain to be invited, but the difficult question is at what level. “Of course you invite them. The question is, what do you invite them to do?” they said. “Can you get to the level of confidence that they will do it in something approaching good faith?
“The only thing worse than them not being there is them being there and derailing it.”
Sunak has placed a lot of political capital in the summit being a success, meaning he will want an international agreement at the end of it on how countries should address AI safety concerns. Ideally, that would include China, but an agreement without it would be better than none at all.
The precedent of the Vienna-based IAEA, which was established in 1957, offers some hope. The watchdog was set up in the early years of the nuclear weapons age and required co-operation between the US and Soviet Union despite their many differences.
While an organisation to regulate AI is unlikely to be established at this summit, the Prime Minister will be hoping to at least lay the groundwork.
A government spokesman said: “The UK is bringing together global leaders and experts for the world’s first major global summit on AI safety in November – driving targeted, rapid international action on the guardrails needed to support innovation while tackling risks and avoiding harms.”
The spokesman said that the Government does not speculate on invitees ahead of summits.
