Analysis: Anthropic and OpenAI consecutively reveal security vulnerabilities, raising concerns about AI model safety

Mars Finance News, according to The Information, reports that Anthropic and OpenAI have experienced security incidents, sparking market concerns about the safety of AI models themselves. Currently, Anthropic is investigating a possible unauthorized access to its Claude Mythos model by users. Almost simultaneously, OpenAI was also exposed for accidentally opening multiple unreleased models in its Codex application. Industry experts point out that these vulnerabilities highlight the need to scrutinize the security governance capabilities of AI companies and also reflect that, during the rapid development of AI technology, the security system still needs improvement. Analysts believe that even AI model providers focused on cybersecurity capabilities still face significant security challenges. As AI gradually is used for defending against cyberattacks, platform security and access control issues also become critical risk points.