OpenAI, Anthropic to collab with NIST on AI safety testing

The standards agency's AI Safety Institute obtained access to "major models" from the two AI leaders ahead of their public release for testing purposes.