OpenAI 的 GPT-Red 攻击率达 84%：AI 红队测试击败人类
openai 內部的 gpt-red 模型在 84% 的測試中發現安全漏洞，而人類紅隊測試人員僅為 13%。這種自動化強化學習方法降低了提示詞注入風險，儘管仍有 3.8% 的攻擊能成功。
openai trained gpt-red using self-play reinforcement learning to simulate malicious prompt injections, finding vulnerabilities in 84% of tests versus 13% by human red teamers. this automated process improved gpt-5.6 sol, reducing direct prompt injection failures sixfold compared to previous models. however, robust attacks still penetrate defenses 3.8% of the time, indicating remaining vulnerabilities. openai integrates these findings directly into model training but has not released gpt-red publicly.