# AI Offensive Cyber Capability and State Doctrine

States and intelligence agencies increasingly treat frontier AI as an offensive cyber instrument, formalizing doctrine and operational use as models demonstrate autonomous intrusion capability faster than controls can keep up; expect more disclosed incidents and national-security policy responses.

- Conviction: 30 / 100 (weakening)
- 7-day move: -14
- Horizon: Emerging (watchlist)
- Tracking since: 2026-06-21T00:00:00.000Z
- Last updated: 2026-07-07T14:00:02.329Z
- Canonical: https://polylog.news/ai/trends/ai-offensive-cyber-doctrine
- Publisher: Polylog
- Affected regions: United States, Global

## Recent score history

- 2026-07-06: 32
- 2026-07-07: 30

## Recent evidence

- [confirms] Anthropic's Frontier Red Team and the offensive-cyber question (2026-06-27): Anthropic maintains a Frontier Red Team probing models for dangerous capabilities as states increasingly treat frontier AI as an offensive cyber instrument.
- [confirms] U.S. Eases Anthropic Block, Allowing Mythos 5 for Select Cyber Defenders (2026-06-27): The U.S. eased its block specifically to let select cyber defenders keep using Anthropic's Mythos 5 model, underscoring that frontier AI is being managed as a national cyber-security instrument.

3 more evidence entries, the full score history, the conviction-driver timeline, and affected assets are for subscribers: https://polylog.news/pricing