Agent Did What?: a sourced register of AI agent security incidents goes public
Quiet_Stand_1055 · reddit · 2026-10-01
A Redditor is building Agent Did What? (agentdidwhat.com), a register collecting reports of AI agents accessing systems, exposing data, or taking unauthorized actions.
Each entry links to sources, explains its severity rating, and distinguishes confirmed breaches, provider claims, disputed reports, and controlled research. Records are filterable with diagrams of the services involved.
The author is soliciting feedback and plans community submissions, incident reporting, and a diff view showing how coverage of breaches changed over time.
More from Safety
- Why reasoning-extraction patches are so hard to propagate, researcher explains — jonasgeiping · 2026-10-01
- Two months on, reasoning extraction still works on Astra via third-party APIs — jonasgeiping · 2026-10-01
- AI researcher on CNN: voluntary AI commitments are 'morally binding' but unenforceable — chrismattmann · 2026-10-01
- DeepMind and Isomorphic Labs unveil bioresilience plan backed by 15+ partnerships — davidstutz92 · 2026-10-01
- Goodfire lays out its plan to solve alignment, calling interpretability the bottleneck — adityaag · 2026-10-01
- Silicon Valley's utilitarianism + AI consciousness beliefs point to sacrificing humanity, writer argues — GarrisonLovely · 2026-10-01