Research notes

Long-form notes from the lab. Write-ups cite vendor engineering blogs, papers, spec text, and CVE records — not aggregator content farms. Items lacking a primary source are marked as such and excluded from detection content until confirmed. New notes are announced in the Atom feed.

• AI for adversarial emulation — primer and lab curriculum

  2026-10-23 · research note · 4 min read

  What adversarial emulation is (and isn't), what AI agents change — adaptivity, machine speed, per-generation polymorphism, near-zero marginal cost — and the lab's curriculum for emulating machine-speed adversaries against its own behavioral Sigma rules.

• AI-Agentic Threat Landscape — curated brief

  2026-10-22 · research note · 6 min read

  A primary-source survey of AI-agent threats as of October 2026: the Poppy Personal Agent Protocol attack surface, documented agentic attack teams (GTG-1002, DUSTMAKER/UNC6780), and autonomous-offense milestones, with the Project Black addendum on AI-driven polymorphic tradecraft.