Hacker News new | ask | show | jobs
Continuously hardening ChatGPT Atlas against prompt injection attacks (openai.com)
3 points by forks 161 days ago