A day after winning China’s support for a worldwide initiative to handle the hazards of the technology, British Prime Minister Rishi Sunak stated on Thursday that world leaders had an obligation to confront the perils of artificial intelligence.
Governments and organizations are racing to create regulations and protections after some tech and political leaders warned that if AI research is not reined in, it might become an existential danger.
At Bletchley Park, the site of Britain’s World War II code-breakers, Sunak conducted discussions on the last day of the inaugural AI Safety Summit on Thursday. Later, he spoke with tech billionaire Elon Musk in central London.
Presidency of the European Commission Ursula von der Leyen, Vice President of the United States Kamala Harris, Secretary-General of the United Nations António Guterres, and other dignitaries were scheduled to confer on a strategy for government-supported testing and assessment of AI models before their public release, according to Sunak’s office.
“I wanted us to have a session to talk about this issue as leaders with shared values in private,” Sunak said at the beginning of the meeting. He said they had a “responsibility to address” threats ranging from extreme perils from artificial intelligence to societal damage and misinformation.
In five years, Sunak hoped they would reflect on this time and realize that they had made the right decision in utilizing AI to its fullest potential while also ensuring our communities’ safety. A subsequent session was scheduled to include representatives from businesses including Google DeepMind, Microsoft (MSFT.O), xAI, Anthropic, Microsoft-backed OpenAI, and Microsoft (META.O).
The transcript of the EU’s von der Leyen speech states that she stated the need for globally recognized AI safety standards to be created. Complex algorithms, according to her, could never be thoroughly tested; thus, “we must ensure that developers act promptly when issues arise, both before and following the release of their models onto the market.”
An interview between Musk and Sunak, which is scheduled to air later on Thursday on Musk’s X, the website that was once known as Twitter, will serve as the last remarks on AI of the two days.
Two insiders at the meeting said that on Wednesday, Musk advised other delegates not to hurry into passing AI laws.
Instead, he said that businesses employing the technology were better positioned to identify issues and could report back to the legislators who were drafting new legislation.
On Wednesday, over one hundred government representatives, academics, and tech industry executives talked about the dangers of misusing AI, the unpredictability of its advancements, control loss, and other end-of-the-world possibilities.
A Chinese vice minister joined tech executives, presidents of the United States and the European Union, and others at the meeting on compelling general-purpose models known as “frontier AI” on Wednesday, marking a first for Western attempts to oversee its safe growth.
More than 25 nations, including the US, China, and the EU, signed the “Bletchley Declaration” on Wednesday, stating that countries must cooperate and develop a shared monitoring strategy.
It focuses on identifying dangers that are of common concern, advancing scientific knowledge of them, and creating international strategies to reduce them.
To prevent regulation from impeding AI businesses from realizing their full potential, governments are currently working with them to map out the future.
Given the lack of confidence in Chinese participation in technology between Beijing, Washington, and many European cities, some British legislators have questioned whether China should be present at the summit.
However, according to Sunak, a discussion about artificial intelligence would not be complete without including one of the unquestionable global leaders in the field.
“I think it’s a good sign of progress that we, the Americans, the Europeans, or China, have all signed the same piece of paper,” he said on the Politico news website’s Power Play podcast.
