The Joe Biden Presidency-led US administration and Rishi Sunak-led British government will be working together to standardise AI testing, and ensure that AI models are safe for the public. Both countries will also establish government-led AI safety institutes.
The United States and Britain have announced a new partnership in which they will be working together to address concerns about the safety of artificial intelligence (AI) technologies, as reported by Reuters. US Commerce Secretary Gina Raimondo and British Technology Secretary Michelle Donelan signed a memorandum of understanding in Washington to collaborate on the development of an advanced AI model.
This partnership follows commitments made at an AI Safety Summit in Bletchley Park in November.
Raimondo emphasized the importance of AI, calling it the defining technology of our generation. She stated that the partnership would accelerate efforts to address national security risks and broader societal concerns associated with AI.
Both countries are establishing government-led AI safety institutes. Britain announced its institute in October to examine and test new types of AI, while the United States launched its own safety institute in November to evaluate risks from advanced AI models. The US institute is now collaborating with 200 companies and entities.
Under the formal partnership, Britain and the United States plan to conduct joint testing exercises on publicly accessible AI models. The two countries are also considering personnel exchanges between the institutes and that have similar partnerships with other countries to promote AI safety.
Generative AI, which can create text, photos, and videos in response to open-ended prompts, has raised excitement as well as concerns about job displacement, electoral interference, and potential catastrophic effects.
In a joint interview with Reuters, Raimondo and Donelan stressed the urgency of addressing AI risks. Donelan highlighted the need for joint action as more capable AI models are about to be released. Raimondo announced plans to discuss AI issues at a meeting of the US-EU Trade and Technology Council in Belgium.
The Biden administration intends to announce additions to its AI team soon, pulling in resources from across the US government. Both countries plan to share information on AI capabilities and risks and collaborate on technical research.
In October, President Biden signed an executive order aimed at reducing AI risks, and in January, the Commerce Department proposed requiring US cloud companies to monitor foreign access to US data centers for AI training. Britain announced plans in February to invest over £100 million to establish new research hubs and train regulators on AI technology.
Raimondo expressed particular concern about the potential misuse of AI for bioterrorism or nuclear war simulation, emphasizing the need for zero tolerance for such capabilities.
