AI start-up Anthropic said it has begun restoring customer access to its latest Mythos 5 and Fable 5 models, after the US Commerce Department lifted security restrictions.
The company said it has increased safeguards to allay government fears that the models could be used to automatically detect software vulnerabilities in critical systems and implement complex hacks.
“Researchers from the US Department of Commerce’s Center for AI Standards and Innovation (CAISI) have tested both our prior and new safeguards and agree that they are extraordinarily strong,” Anthropic stated.
Image credit: Shutterstock
Security risk
