AI developer toolsBriefChecked against a primary source
AI-orchestrated and agentic attacks in 2026
Anthropic, Sep 2026: “automated exploit foundries”.
- Disclosed
What happened
- Anthropic, Sep 2026: “automated exploit foundries”. GTG-10007 got “more than a dozen possible zero day findings in a single month” against network appliances. Prompt injection was used against LiteLLM wrapper services to steal keys.
- CrowdStrike, 28 Oct 2026: four actors between May and Aug 2026, two with advanced AI harnesses. All bypassed model safeguards.
- Google GTIG, Q2 2026: an agent-enabled mass credential harvesting campaign was planned and executed in under 6 hours.
Sources
- https://www.anthropic.com/threat-intelligence-report-september-2026
- https://cloud.google.com/blog/topics/threat-intelligence/from-prompting-to-autonomy-the-evolution-of-adversarial-ai
- https://www.cybersecuritydive.com/news/agentic-ai-cyberattacks-harnesses-crowdstrike/833714/
This is a brief: a short, sourced summary from our incident tracking. Full advisories add commands to check whether you are affected.