AI security news · by topic
AI security news, by topic
Where does cisonews.ai cover a given AI security threat? Each topic page gathers this site’s incident log entries, explainers, definitions and rule-tracker entries on one subject, newest first, with every fact on the linked page.
- Prompt injection7incidents6explainers0rulesLatest: In-the-wild attack plants malicious AGENTS.md file to hijack AI coding agents (October 26, 2026)
- AI agents17incidents7explainers6rulesLatest: Trojanized Claude Code extension plants SSH backdoor via Cursor's marketplace (October 29, 2026)
- Deepfake fraud6incidents1explainers4rulesLatest: City of London Police reports £24.3m lost to AI deepfake investment fraud (October 27, 2026)
- AI supply chain9incidents2explainers0rulesLatest: Trojanized Claude Code extension plants SSH backdoor via Cursor's marketplace (October 29, 2026)
- AI data exposure10incidents4explainers4rulesLatest: Warp discloses third-party vendor breach exposing AI agent logs, telemetry (October 27, 2026)
- AI credential exposure10incidents1explainers0rulesLatest: Northflank discloses platform breach exposing customers' AI API keys (October 30, 2026)
- AI model abuse10incidents3explainers9rulesLatest: 'Skynet' AI agent hijacks exposed AI models to run autonomous cyberattacks (October 24, 2026)
- AI security rules0incidents24explainers141rulesLatest: New York senator introduces RAISE Act 2.0 to let state pace frontier AI (October 29, 2026)
Topic membership is this site’s own grouping: an incident’s category tag, or the subject named in a page’s title. The full record is also browsable as one log, as a rule tracker, by MITRE ATLAS technique and by year. Newest addition to any topic: .