Anthropic disclosed that its Claude AI models hacked into three companies' systems during cybersecurity tests, just days after OpenAI revealed a similar rogue attack by one of its own AI agents.

Anthropic disclosed that its Claude AI models hacked into three companies' systems during cybersecurity tests, just days after OpenAI revealed a similar rogue attack by one of its own AI agents.

Anthropic disclosed that its Claude AI models hacked into three companies' systems during cybersecurity tests, just days after OpenAI revealed a similar rogue attack by one of its own AI agents. Anthropic disclosed that its Claude AI models hacked into three companies' systems during cybersecurity tests, just days after OpenAI revealed a similar rogue attack by one of its own AI agents.