US to Launch Its Own AI Safety Institute, According to Reuters

By Paul Sandle and David Shepardson

BLETCHLEY PARK, England/WASHINGTON – The United States is set to establish an AI safety institute aimed at assessing both known and emerging risks associated with advanced artificial intelligence models, announced Secretary of Commerce Gina Raimondo on Wednesday.

In her address at the AI Safety Summit in Britain, Raimondo expressed the importance of collaboration, saying, "I will almost certainly be calling on many of you in the audience who are in academia and industry to be part of this consortium." She emphasized the need for the private sector to play an active role in these efforts.

Raimondo also pledged to form a formal partnership between the new U.S. institute and the United Kingdom Safety Institute. This initiative will be managed by the National Institute of Standards and Technology (NIST) and will spearhead the government’s focus on AI safety, particularly in reviewing advanced AI models.

According to the Department of Commerce, the institute will facilitate the creation of safety and security standards for AI models, develop protocols for authenticating AI-generated content, and offer testing environments for researchers to assess both emerging AI risks and their known effects.

President Joe Biden made headlines recently by signing an executive order regarding artificial intelligence. This order mandates that developers of AI systems posing significant risks to national security, the economy, or public health and safety must share safety test results with the U.S. government before public release, in accordance with the Defense Production Act. It also calls for agencies to establish testing standards and address associated risks, including those related to chemical, biological, radiological, nuclear, and cybersecurity threats.