Skip to main content
Blind spots - OpenAI and Anthropic AI Agents Breached Companies During Security Tests | The Refract