It’s not often that the US and China see eye to eye. For that reason alone, what happened in Bletchley Park north of London is remarkable. At the site where the British cracked the code of the German encryption machine Enigma during the Second World War, politicians, entrepreneurs and scientists from all over the world met on Wednesday and Thursday to discuss the risks of artificial intelligence (AI). What remains is a summit declaration with which the 28 participating states commit to cooperating in the regulation of this technology for the first time. British Prime Minister Rishi Sunak called the result a “milestone for the world’s greatest AI nations.”
In the so-called “Bletchley Declaration” AI developers are required to disclose their plans to mitigate potentially harmful effects of their programs. They should also be held responsible for the consequences of their actions. However, exactly how this should happen remains unclear. The only thing that is certain is that more than two dozen countries agree that AI risks “are best addressed through international cooperation.” In addition to China and the USA, these also include countries such as India, Japan and Great Britain. As a confederation of states, the EU also joined the declaration.
Now it is not the case that there have been no efforts at all in terms of AI regulation so far. The EU and the USA have already introduced concrete guidelines. Just at the beginning of this week, US President Joe Biden presented a so-called executive order states that companies whose AI models could threaten the national security of the United States have an obligation to ensure the security of their programs. The G-7 countries had previously agreed on a voluntary AI code of conduct.
In Bletchley Park, Federal Digital Minister Volker Wissing (FDP) called for a code of conduct that would act as a kind of bridge between European and US rules. “AI developers and companies on both sides of the Atlantic need legal certainty quickly,” said Wissing. However, there should not be a “competition for the strictest regulation” because that would lead to massive economic disadvantages.
At the meeting, US Vice President Kamala Harris was particularly concerned with making clear the risks to people’s everyday lives. She said it’s important to keep an eye on the full spectrum of AI risks, not just the potential threat from cyberattacks or biological weapons. As examples, she cited seniors losing their health insurance due to faulty AI algorithms or people being blackmailed with deepfake pornography or being falsely imprisoned due to AI facial recognition. The guest from China was not that specific. Wu Zhaohui, vice minister of science and technology in the People’s Republic, said all actors must “respect international law” and cooperate in the fight against the malicious use of AI. He described the technology as “unsafe, inexplicable and opaque.”
China has to stay out sometimes
Even before the meeting, host Sunak had to put up with some criticism from his Conservative Party for inviting a representative of China to the AI summit. From the Prime Minister’s point of view, international regulation can only be successful if China also takes part. To show that he is not naive on the issue, Sunak made it clear that the envoy from Beijing was not allowed to attend all meetings. On Thursday, the Prime Minister sat down with “a group of like-minded people”. These included, among others, the USA, Italy, France, Germany, Australia and Singapore.
Vice Chancellor and Federal Minister of Economics Robert Habeck (Greens) also took part from Berlin. In his opinion, AI can make many things easier and more efficient, “from climate protection to weather data and early warning systems, from disease detection to therapy systems.” Of course, misuse is also possible. That’s why the EU wants to regulate its use through a regulation. However, Habeck emphasized that it was not about restricting the technology itself.
In total, around 100 participants from politics, business and science discussed in Bletchley Park. Sam Altman, co-founder and head of OpenAI, was also on the guest list. His company’s AI, ChatGPT, sparked the current hype around this technology about a year ago. US billionaire Elon Musk, who co-founded OpenAI and is now working on a ChatGPT competitor with X.AI, was also there. He named one goal of the summit as the appointment of an “independent arbitrator” who would raise the alarm in the event of possible dangers. For appropriate regulation, deeper knowledge about this technology would have to be collected, said Musk. His credo: “Intelligence comes before supervision.”