Detect Deepfakesby Resemble AI
The Incident Database

Deepfake and agentic AI incidents we could verify.

A curated, cited database of deepfake incidents and agentic AI attacks — from voice-clone fraud to real-world abuse of web and computer-use agents. Agentic entries cover actual security incidents and scams, with evidence of an AI agent’s role. Research demonstrations and hypothetical threats are excluded. Updated daily.

5053
Documented incidents
679
Audio / voice
1303
Video / face-swap
1617
Image / synthetic
116
Agentic AI attacks

2026· 60 incidents

November 2026

·60

Browsing 1,773 of 5,053 documented incidents. Entries involving minors, non-consensual sexual imagery, or private individuals are retained in the database for research and enforcement use but are not published as browsable pages. Researchers who need the full record can get in touch.

Cite this database

Journalists and researchers: cite us freely.

The Deepfake & Agentic AI Incident Database is a public, continuously-updated resource maintained by Resemble AI. Every entry has primary-source citations. Link directly to incident pages — we keep URLs stable.

Suggested citation
Resemble AI, “Deepfake & Agentic AI Incident Database,”
https://www.detectdeepfakes.com/examples, accessed Nov 2026.