Rishi Sunak wants to position the UK as a global leader in coordinating regulations and international action on AI. Photo: Ian Fogler/AFP
World leaders will warn of potential “catastrophic harm” caused by artificial intelligence (AI) at a global summit in the UK next month.
Draft summit communique, which must be signed by participants, warns that artificial intelligence systems could be used to carry out cyberattacks and create biological weapons, and that it is “especially urgent” to address these risks.
The statement, which is being discussed among participants and may be modified in the coming weeks, also argues that artificial intelligence systems may try to “increase their own influence and reduce human control.”
Wartime Codebreaking Summit Center at Bletchley Park aims to provide a unified global response to the emergence of increasingly powerful artificial intelligence systems.
The draft agreement does not propose new laws or a global regulator.
Rishi Sunak, the prime minister, aims position Britain as a global leader in coordinating regulations and international action on technology.
The two-page communiqué, seen by The Telegraph, said AI “represents a huge global opportunity” but warned it “also poses significant risks”, citing concerns in particular that it could be used to hack or development of biological weapons.
The draft statement says: “The most significant of these risks arise from potential intentional misuse or control issues, where AI systems may seek to increase their own influence and reduce human control, and these problems are partly due to the fact that these capabilities are not fully understood.
“We are particularly concerned about such risks in areas such as cybersecurity and biotechnology. The most dangerous capabilities of these AI models could result in significant, even catastrophic harm, both intentional and unintentional.
“Given the rapid and uncertain rate of change in AI, and in the context of accelerating technology investments, we affirm, that deepening our understanding of these potential risks and actions to address them is particularly urgent.”
It is understood that some of the draft's harshest language does not reflect the government's position and is likely to be withdrawn.
The communiqué is due to be finalized next week and signed by governments on November 1, the first day of the two-day summit. In recent weeks, officials have been busy discussing its formulation with the US, EU, G7 countries and China.
In a sign of respect for China, whose invitation to the summit has been controversial, the statement said countries should “engage a wide range of partners as appropriate.” China is expected to attend at least part of the summit, although its presence on the security-focused second day of the event remains unclear.
The draft commits to developing an “international scientific network” for AI safety research and assistance. develop laws around this technology. It pledges to “strengthen and maintain our collaboration to identify, understand and take appropriate action” on risks, including through future summits.
It also calls on major artificial intelligence labs to invest in security testing tools. < /p>
He is not proposing the creation of a global regulator, although UK officials are likely to propose that the government's £100 million AI taskforce take a leading role in research into the technology and become a permanent “AI Safety Institute”.
Concerns about artificial intelligence have increased over the past year due to advances in technologies such as ChatGPT, developed by Microsoft-backed startup OpenAI. The technology has also led to warnings of election interference, a topic scheduled to be discussed at the summit.
However, some experts said the government is exaggerating the risks posed by AI.
Source close to Michelle Donelan, Minister for Technology, said: “Michelle has never shied away from the risks associated with the development of artificial intelligence, but she is equally inspired by the transformative opportunities that can be unlocked through AI.”
< p>“We must find a way to manage risk to safely pursue these opportunities. That's what we hope to start talking about at the summit next month.”
Recent Comments