Designing AI agents to resist prompt injection

kept by eddie mar 12

OpenAI argues that defending against prompt injection in AI agents requires treating attacks as social engineering and designing systems to constrain impact rather than just filtering inputs.

A backdoor in a LinkedIn job offer - Roman Imankulov
A fake recruiter sent a LinkedIn job candidate a GitHub repo with a backdoor that executes on npm install by running a remote-controlled command payload hidden in a test file.
also on #social-engineering, #security
Claude Cowork exfiltrates files
Researchers demonstrate file exfiltration from Claude Cowork via prompt injection, exploiting an acknowledged but unremediated vulnerability.
also on #prompt-injection, #ai-safety, #security, #ai-agents
Keeping your data safe when an AI agent clicks a link
OpenAI prevents URL-based data exfiltration by only allowing AI agents to auto-fetch URLs that were independently observed in a public web index.
also on #prompt-injection, ChatGPT, #security, #ai-agents