Researchers use Anthropic models to breach OpenAI systems, exposing AI security gaps

Cybersecurity researchers leveraged Anthropic's AI models to infiltrate OpenAI's systems, revealing vulnerabilities in the ChatGPT maker's defenses. The incident underscores growing security scrutiny on leading AI firms, though details of the attack and impact remain undisclosed. This marks a notable cross-competitor exploitation, highlighting systemic risks in AI supply chains.