NIST’s AI Safety Institute signs agreements with Anthropic, OpenAI on pre-deployment testing

The U.S. AI Safety Institute has announced first-of-a-kind agreements with two major artificial intelligence companies that will provide AISI with access to “major new models” for evaluating safety risks both before and after deployment.

“Today, the U.S. Artificial Intelligence Safety Institute at the U.S. Department of Commerce’s National Institute of Standards and Technology announced agreements that enable formal collaboration on AI safety research, testing and evaluation with both Anthropic and OpenAI,” NIST said on Aug. 29.

“Each company’s Memorandum of...