Anthropic Blocks Access to AI Models Amid Alleged Unauthorized Chinese Access

Anthropic reportedly blocked access to its AI models, including Mythos, following allegations that China may have gained unauthorized access to the system. The restriction is linked not only to a claimed jailbreak but also to concerns over foreign access, as stated by an anonymous source. No specific technical details, dates, or methods of compromise were disclosed in the report. The incident involves Fable and Mythos, two AI models developed by Anthropic, though their exact roles or vulnerabilities remain unspecified. The article does not clarify whether the access was exploited for data extraction, manipulation, or other malicious purposes.