Anthropic’s Fable 5 Safeguards Were Always A ‘Judgement Call’
Anthropic has unexpectedly disabled its Fable 5 model following a U.S. government directive citing national security concerns related to a potential jailbreak. The company stated it disagrees with the directive, arguing that a narrow vulnerability should not warrant recalling a widely deployed model, and believes such a standard would halt industry-wide deployments. White House AI czar David Sacks criticized Anthropic for prioritizing the consumer offering over safety, while other tech executives reportedly raised similar concerns to the government. AI
IMPACT Government intervention in AI model releases could set a precedent for future deployments and safety standards.