China, the United States, the European Union and about 20 countries in Britain on Wednesday signed the Bleachley Declaration for the "safe" development of artificial intelligence, during the first international summit on "Artificial Intelligence Security and Safety", which examines the rapid development of this technology.
British Prime Minister Rishi Sunak welcomed the announcement via X, saying that "this historic announcement marks the beginning of a new global effort to build public trust in AI by ensuring its safety."
The European Union and 28 countries, representatives of which met at Bletchley Park, north of London, agreed on the "urgent need to collectively understand and manage the potential risks" of AI through a "new global effort to ensure that AI is developed and deployed in a safe and responsible manner".
Faced with the growing potential of software like ChatGPT, Bletcheley's announcement "shows that for the first time the world is coming together to identify the problem and highlight related opportunities," British Technology Minister Michelle Donnellan told AFP.
The meeting, which began on Wednesday and concludes on Thursday, "is not aimed at laying the foundations for global legislation, but must chart a way forward" in this area, she said.
She added from this highly symbolic site, which housed the Center for Decoding in World War II, that this summit will be followed by two international summits on artificial intelligence, the first in South Korea in 6 months, and the second in France in a year.
Over the course of the two days of the summit, political leaders, AI experts and tech giants will gather at the initiative of the United Kingdom, which wants to take the lead in global cooperation in this technology.
Meanwhile, US Vice President Kamala Harris, in a speech in London, is scheduled to announce the establishment of an artificial intelligence security institute in Washington. The institute, similar to the one the UK also announced, will bring together experts and those responsible for developing "guidelines" and evaluating the most advanced AI models to "identify and mitigate risk," according to the White House.
Generative AI, capable of producing text, sounds or images on a simple request in a matter of seconds, has made impressive progress in recent years, and new generations of these models are expected to emerge in the coming months.
The British government warned in a report published on Thursday that the technology, which raises high hopes for medicine or education, could also pose an "existential threat" by destabilizing societies by making it possible to manufacture weapons or evade human control.
Rishi Sunak welcomed the announcement of the signing of the Bletchley Declaration Agreement for the Secure Development of Artificial Intelligence (Anatolia)
The presence of politicians on the second day
After the first day of the summit devoted to talking about the potential risks of more advanced artificial intelligence, high-level political representatives are expected to attend Bleachley Park on Thursday.
They include European Commission President Ursula von der Leyen, UN Secretary-General António Guterres, and Italian Prime Minister Giorgia Meloni, the only Group of Seven leaders to attend the summit.
The British government hopes to at least reach the first international declaration on the nature of AI risks. It also proposes the creation of a group of AI experts based on the model of the Intergovernmental Panel on Climate Change (IPCC), which is responsible for monitoring climate issues.
Billionaire Elon Musk, head of Network X, is holding talks with British Prime Minister Rishi Sunak after the summit. Musk on Wednesday stressed the need for "governance" in artificial intelligence, saying the technology poses one of the "biggest threats" facing humanity.
"What we aim to do is create a framework for better understanding, so that there is at least an independent judge who can monitor what AI companies are doing and sound the alarm if he has concerns," Musk told reporters at the summit. "I hope at this summit there will be an international consensus on the initial understanding of advanced AI."
Musk stressed need for 'referee' in AI (Getty Images)
In an open letter published on Tuesday, many of the technology's "founding fathers", such as Yoshua Bengeo and Jeffrey Henton, called for "the development and ratification of an international treaty on artificial intelligence" to reduce the risks and "potentially catastrophic consequences" that these systems impose on humanity.
The challenge that countries face in dealing with AI lies in their ability to ensure that AI works in line with global standards without hindering or restricting the innovation of AI labs, i.e. regulating the work of this technology, as the European Union and the United States have chosen.
Last week, companies such as OpenAI, Meta, owner of Facebook and Instagram, and Google, owner of DeepMind, agreed to announce some of their AI security rules at the UK's request.
In an open letter to Rishi Sunak, some 100 international organizations, experts and activists expressed dismay that the summit was held "closed", with tech giants dominating it and a limited civil society presence.