Why Anthropic's 'safe' Mythos-class model won't answer questions about cancer

• Anthropic’s Claude Fable 5 model has advanced safeguards that can be triggered by mundane requests.
• The Mythos-class model is blocked or reverts to Opus 4.8 when flagged by safety classifiers.
• Anthropic said the safeguards were necessary to release the model to the public.

“That gap in understanding could be really dangerous for causing policymakers, or for that matter the public, to not fully understand the risks that these models pose in terms of the capabilities they offer,” he said.