新闻
Responding to the next frontier of critical cyber capabilities
OpenAI · 发布于 · 阅读约3分钟
30秒读懂
- 发生了什么
- OpenAI released preliminary cybersecurity evaluations for Astra and described safeguards and security control improvements.
- 为何重要
- Security teams and infrastructure engineers should care when assessing AI model deployment risks and compliance requirements.
- 注意
- These are preliminary evaluations, not final security certifications. Full scope of Astra's capabilities and limitations remains unclear.
收听本摘要
- eval
The Agent Architect
每周一个模式、一个权衡、一个生产事故案例。为构建智能体系统的人准备的每周简报。
每周一封邮件,一键退订。您的地址仅用于发送简报。