External · Stanford UIT · beginner

Responsible AI at Stanford

The Information Security Office's guide to using generative AI without exposing Stanford data: be aware that third-party tools store what you paste, be careful (no moderate or high-risk data, PHI, passwords, source code or NDA material in tools without a Stanford agreement), be transparent about AI in your work, and expect hallucinated facts and citations. Includes best practices by area and links to the approved services list.