1 articles about adversarial testing
OpenAI has published GPT-Red, a safety research model that autonomously probes and hardens other AI systems through self-improvement loops and adversarial testing.
