Michelle Bowman said regulators must assess how systems such as Anthropic’s Mythos could improve cybersecurity or be misused to find weaknesses. Anthropic has limited the model’s release while officials review safeguards.
verifying reliability
No specialized terms available for this topic.