19:25
0
OpenAI 在员工警告代码访问权限过于宽泛后加强安全措施
According to reports, OpenAI employees warned management before ChatGPT's launch that overly broad code and research access permissions could expose sensitive technologies. The internal risks included excessive permissions to code repositories and research materials widely stored on GitHub and Slack. OpenAI has since strengthened its security practices and plans to launch a new security initiative to address these vulnerabilities.
展开