OpenAI Breached by Researchers Using Anthropic Models

This breach highlights security vulnerabilities in AI systems where external models can compromise providers, raising concerns about AI safety and competitive integrity.

Auto-scored: The breach constitutes a real escalation in AI security risks due to unauthorized access.