OpenAI News·· 2025-12-22精选AI 评分65
OpenAI 加强 ChatGPT Atlas 对提示注入攻击的防御
Continuously hardening ChatGPT Atlas against prompt injection
AI 导读
OpenAI 正在使用强化学习训练的自动化红队技术来加强 ChatGPT Atlas 对提示注入攻击的防御，以提前发现新型漏洞并加固浏览器代理的防御能力。
推荐理由
OpenAI 正在加强 ChatGPT Atlas 对提示注入攻击的防御，读者可以关注这一进展对 AI 安全领域的影响。
原文
OpenAI is strengthening ChatGPT Atlas against prompt injection attacks using automated red teaming trained with reinforcement learning. This proactive discover-and-patch loop helps identify novel exploits early and harden the browser agent’s defenses as AI becomes more agentic.
来源：OpenAI News · openai.com