The UK hosted an AI Safety Summit this week, drawing attention not only to progress in AI safety, but also to growing international divisions over the future of AI and its governance.
Analyst Rory Bathgate, from the British magazine IT Pro, wrote an interesting article analyzing the event. According to Bathgate, although the summit was attended by representatives from the world’s leading AI nations, including the US, UK, China, France and Japan, among others, it was the underlying dynamic that dominated the discourse. These countries came together to sign the Bletchley Declaration, uniting 28 countries under the umbrella of shared risks and opportunities of advanced artificial intelligence. They pledged to strengthen international cooperation in research and climate change mitigation, an action that at first glance appeared to unite the world’s AI leaders.
However, the nuances hidden in the statement, unspoken and perhaps more fundamental, are likely to be the driving force in the field of artificial intelligence over the coming year. The signatories expressed their support for a comprehensive international network focused on AI border security. However, the actions of some countries seem to point to unilateral agendas. For example, the United States, represented by Commerce Secretary Gina Raimondo, used the summit as a platform to announce a new US Institute for AI Safety. This followed President Joe Biden’s executive order setting stringent requirements for AI developers, reinforcing the idea that the US was using the summit to advance its national interests in AI.
The United Kingdom was not without national differences either. In the middle of the summit, technology giants HPE and Dell revealed their plans for new supercomputers. Funded by the British government, these supercomputers will support the UK’s Artificial Intelligence Research Resources (AIRR). Michelle Donelan, Minister for Innovation and Technology, said: “This investment seeks to provide world-class UK researchers with advanced tools, ensuring the UK maintains its leading position in AI safety.
But despite everything the summit witnessed, it had moments that sparked controversy and raised eyebrows. Prime Minister Rishi Sunak’s live conversation with tech mogul Elon Musk seemed like a PR exercise, talking more about pop culture than substantive political issues.
The Chinese perspective, expressed by Yi Zeng of the Chinese Academy of Sciences, emphasized the urgent need to establish an international AI safety network. Even after Bletchley’s announcement, Zeng’s emphasis on collective action to address unexpected advances in artificial intelligence signaled a possible lack of confidence in the strength of the summit agreements.
The British government praised the participation of Wu Zhaohui, Chinese Vice Minister of Technology, as evidence of the comprehensiveness of the summit. Wu’s speech made China’s position clear: he called for expanding the representation of developing countries in AI governance and urged global cooperation to democratize AI knowledge, with an emphasis on open source solutions.
However, as Bathgate noted, at the conclusion of the summit there were no concrete plans for how AI safety research would be shared collaboratively. Despite the general consensus on the importance of such security, it appears that countries will continue to follow their pre-existing paths when it comes to AI legislation and associated regulations.
In light of the strong regulations already in place, such as EU law on AI and the G7 Guidelines on AI, the impact of the Bletchley Declaration remains uncertain. Open source AI, despite its supporters, remains a controversial topic.
Therefore, although the UK AI Security Summit generated many critical discussions and served as a stage for international cooperation, the underlying undercurrents reflected a world in which nations, although appreciative of global cooperation, are intensely protective and competitive over advances and strategies in intelligence. Artificial. Only time will tell how these dynamics will shape the future of global AI.
Illustration: Diario TI via Midjourney
Join the conversation
“Subtly charming bacon junkie. Infuriatingly humble beer trailblazer. Introvert. Evil reader. Hipster-friendly creator.”