The encryption community is concerned that the security mechanism of Claude Fable 5 under Anthropic may be exploited by hackers On June 10th, according to Cointelegraph, Anthropic has released the first public version of the Claude Mythos model, Fable 5, despite the built-in security barrier. Crypto users are concerned that it may be used for malicious purposes. The model has previously identified over 10000 high-risk vulnerabilities in 'systemically important software'. Anthropic stated that Fable 5 is securely prepared for general use, and the fence will redirect topics such as cybersecurity to other models. Moonrock Capital founder Simon Dedic has warned that the cost and skill requirements of using Fable 5 to find vulnerabilities in smart contracts will be "reduced to almost zero," and unaudited protocols will become targets, with known vulnerabilities being continuously repeated. He suggested that users revoke wallet authorization, remove assets from the agreement, and transfer them to a new hardware wallet. Michael Egorov, co-founder of Curve, believes that the threat has been exaggerated, with only a few thousand lines of smart contract code, and that humans and conventional AI can fully reason. The real risk lies in operational security and supply chain attacks.