TECHNOLOGYJun 11 · 07:18 UTCBUSINESS INSIDERShubhangi Goel ([email protected])

Anthropic: 'We made the wrong tradeoff' in new model guardrails

Anthropic reversed a policy on Claude Fable 5's secret safeguards after backlash, now visibly rerouting flagged requests to Opus 4.8. The company admitted making the wrong tradeoff in balancing safety and transparency. Anthropic's Mythos model remains a national security concern due to its advanced capabilities.

Nexus surfaces and summarizes. The full story lives at the source.

Mentioned

AnthropicorganizationClaude Fable 5topicMythostopicOpus 4.8topicnational securitytopic

Spot something wrong with this article?Report a problem →

Forward this

Related Signal

Adjacent reporting

TECHNOLOGYMay 27 · 21:09 UTC
NYC Chancellor Kamar Samuels pledges stronger AI guardrails: ‘We missed the mark’
TECHNOLOGYJun 9 · 21:36 UTC
Anthropic, which claimed AI model was too risky for public to use, releases ‘safe’ version
TECHNOLOGYJun 9 · 17:08 UTC
Anthropic releases guardrailed version of Mythos for public use
POLITICSApr 17 · 20:14 UTC
Anthropic’s new cybersecurity model could get it back in the government’s good graces