UK says it will again test OpenAI and Anthropic models before release

The UK government said its AI Security Institute will keep testing the most powerful AI models before release, reversing a US move that put White House checks first. Anthropic had not shared its Mythos model with the institute, Bloomberg reported.

• UK's AI Security Institute will retain pre-release access to OpenAI and Anthropic models
• The White House had asked firms to hold new models back from UK testers
• Anthropic did not share its Mythos model with the institute
• AI minister Kanishka Narayan set up a task force to guard UK systems from AI attacks