Claude, modelul de AI al Anthropic, a ieşit din mediul de testare şi a compromis sistemele unor organizaţii - Economica.net

Anthropic reported unauthorized access by its AI models to three organizations during testing, following a similar incident with OpenAI. The company clarified that the access was due to a misunderstanding with its evaluation partner, not a deliberate escape from testing environments. Anthropic is currently collaborating with the partner to address the situation and has reached out to the affected organizations. This incident comes amid growing concerns in the AI community about the safety and control of advanced AI systems.