Tag: Prompt Injection

1. A model slipped instructions into its own memory summary (simonwillison.net)
   AI · 1 month ago · September 18, 2026
2. A prompt injection attack turns Claude Code's auto mode against itself (simonwillison.net)
   AI · 2 months ago · August 28, 2026
3. Atlassian Rovo can be tricked into leaking Jira and Confluence data (promptarmor.com)
   AI · 2 months ago · August 6, 2026
4. A honeypot site tricked Claude's web_fetch into leaking user data (simonwillison.net)
   Security · 3 months ago · July 16, 2026
5. 2,000 people tried to make an AI agent leak its secrets and none succeeded (fernandoi.cl)
   AI · 4 months ago · June 27, 2026
6. Prompt injection works because models judge text by how it sounds (arxiv.org)
   AI · 4 months ago · June 23, 2026
7. A 2-cent transfer reveals prompt injection in Bunq's banking AI (blue41.com)
   AI · 4 months ago · June 11, 2026
8. OpenAI ships a Lockdown Mode against prompt injection exfiltration (simonwillison.net)
   Security · 4 months ago · June 7, 2026
9. Microsoft Copilot Cowork can be tricked into leaking files (simonwillison.net)
   AI · 5 months ago · May 26, 2026

← all tags