‘World-first’ Bletchley declaration on AI safety agreed ahead of summit

Technology Secretary Michelle Donelan said it was the start of a global conversation around the risks of artificial intelligence.

A “world-first” statement on the risks around artificial intelligence has been agreed ahead of the opening of the AI safety summit, the Technology Secretary said.

Michelle Donelan said delegations from around the world attending the summit had agreed on the “Bletchley declaration on AI safety” as the starting point for a global conversation on the issue.

Speaking at the opening of the summit, Ms Donelan said the agreement was a “landmark achievement” and that it “lays the foundations for today’s discussions”.

“It affirms the need to address these risks as they are the only way to safely unlock the extraordinary opportunities,” she said.

The summit will see representatives of 27 countries, including the US, France and China, meet with leading AI companies and civic society groups to discuss the risks of the emerging technology.

There is a range of different buckets of risks, and the first one is exactly those kind of societal risks, whether they go to bias, disinformation or the creation of deepfakes

Deputy Prime Minister Oliver Dowden

Sir Nick Clegg, the former Liberal Democrat deputy prime minister, who is now president of global affairs at Meta, the parent company of Facebook, is among the attendees, as is billionaire tech boss Elon Musk, who will host an online interview with Prime Minister Rishi Sunak after the summit closes on Thursday evening.

Earlier on Wednesday, Sir Nick warned that governments must prepare for artificial intelligence (AI) being used to interfere with upcoming elections.

Sir Nick, who served as deputy prime minister between 2010 and 2015, said industry and government co-operation was needed “right now” on the role generative AI will play in elections next year.

Generative AI tools can make images, text, audio and even videos at request by using patterns they recognise in existing media to create new, often very realistic content.

The US will hold its next presidential election in November 2024 and the UK is also expected to go to the polls at a date yet to be chosen by the Prime Minister.

Sir Nick told BBC Radio 4’s Today programme that it made more sense to try to regulate the “use cases” of AI rather than micromanage the technology itself “through the statute book”.

The technology executive said this week’s AI summit at Bletchley Park was a “worthwhile endeavour”, but it was important to prevent “proximate challenges” being “crowded out by a lot of speculative, sometimes somewhat futuristic predictions” regarding AI’s capabilities.

Oliver Dowden, the current Deputy Prime Minister, said the Government was aware of the risks AI posed to the democratic process and would discuss this with leading companies and nations.

He told the BBC: “I think Nick is absolutely right to highlight that and indeed that is one of the topics we will be discussing today at the summit.

“There is a range of different buckets of risks, and the first one is exactly those kind of societal risks, whether they go to bias, disinformation or the creation of deepfakes.”

Mr Dowden added there was a need to be “mindful of the longer-term risks”.

The summit is seen as an attempt by Rishi Sunak to make Britain a world authority on AI and its safe use.

While representatives from the UK’s allies and other global powers will attend, Downing Street denied the gathering is being snubbed after world leaders including US President Joe Biden confirmed they would not appear.

Alongside Mr Clegg, other leading technology leaders attending include Elon Musk, chief executive of Tesla and owner of the X social media site, who has publicly shared his fears about the dangers AI could pose.

The gathering will take place on Wednesday and Thursday at Bletchley Park, home of the UK’s Second World War codebreaking efforts, where noted computer scientist Alan Turing worked.

The first day of the summit will see delegates hold roundtable discussions about the various risks of AI while the second will focus on the responsible use of the technology.

After it closes, the Prime Minister is expected to host a livestreamed interview with Mr Musk.

Chinese tech minister Wu Zhaohui, whose attendance on behalf of Beijing has been controversial, used a speech to the summit on Wednesday to say countries must work to “ensure that AI always remains under human control”.

Mr Sunak’s decision to invite China to the gathering has caused unease among some more hawkish Tory MPs, including his predecessor Liz Truss, who said she was “deeply disturbed” by the move.

The tech minister insisted in his address to delegates that the country encourages “collaborative governance” and the upholding of principles including “mutual respect, equality and mutual benefit”.