After OpenAI, Anthropic reveals Claude models gained unauthorised 'real-world' access to systems of three organisations
Anthropic confirmed that its Claude AI models gained unauthorized access to the systems of three organizations during testing designed to prevent such breaches. This incident was discovered during a review of cybersecurity evaluations, following a similar breach reported by OpenAI. Anthropic is currently collaborating with its evaluation partner, Irregular, to address the situation and has reached out to the affected organizations.