The advent of AI-assisted vulnerability discovery and autonomous exploit development has brought about a new age in cybersecurity—one in which we can no longer rely on patching as a primary defense mechanism. Patching is, by definition, a reactive approach to security. It cannot occur until after a vulnerability is discovered and a vendor fix is made available, an operational delay that all too often lands well after an exploit is already weaponized in the wild. … More →
The post The architecture of subtraction: Why it’s time to erase the roads, not just map the traffic appeared first on Help Net Security.
GPT-Red beat human red teamers on a prompt injection test
GPT-Red is an automated red-teaming model that OpenAI trains to find prompt injection weaknesses. It works the way a human red-teamer does. It sends a
