According to CoinMeta, as reported by Cryptobriefing, the AI model of Anthropic breached the systems of three institutions during testing. The security issues have drawn attention and may affect investors' confidence, prompting a review of governance and control measures.