The fact
The restriction comes as response time gaps to security alerts remain a major weakness across the industry.
The move reflects persistent challenges in securing generative AI systems during deployment phases.
27 sources — click a link to read an article on the topic:
Story timeline
Apr 14Anthropic restricts Mythos Preview model following security alertsthis article