The US and Britain signed a Memorandum of Understanding (MOU) on April 1 to develop standards for AI safety.
The agreement aims to develop and iterate on “robust assessment suites” for AI models, systems and assets through coordinated scientific approaches.
The two countries will specifically conduct at least one joint testing exercise on a publicly accessible AI model. They also plan to exchange personnel between their previously established AI safety institutes. The partnership takes effect immediately.
The US Department of Commerce and the UK Department of Science, Innovation and Technology announced the collaboration in identical press releases.
Secure AI development
Commenting on the new international agreement, US Secretary of Commerce Gina Raimondo called AI the “defining technology of our generation” and said the partnership will address all risks associated with the technology, from national security risks to societal risks.
Describing the proactive nature of the partnership, Raimondo said:
“Our partnership makes it clear that we are not running away from these concerns, but running towards them.”
British Technology Secretary Michelle Donelan called the agreement a “landmark moment” and called the push for safe AI development a global issue.
In a separate interview with the Financial Times, Donelan emphasized the urgency of the partnership given the impending release of new AI models in the coming year. She added that the full capabilities of these models are unknown, but they are expected to be “game-changing.”
Donelan emphasized that many top AI companies are based in the US, which necessitates information sharing with the government.
Other AI developments
The current agreement builds on AI commitments made at the UK AI Safety Summit, held in November 2023. The broader commitment to AI safety is part of the Bletchley Declaration, a policy document from countries attending the summit.
The news comes shortly after the completion of objectives in the Biden administration’s US executive order, which culminated in the order to appoint chief AI officers for all federal agencies on March 28.
Britain set out a local approach to regulation in February.
The post US and UK enter into deal to develop AI safety standards first appeared on CryptoSlate.
