AI AI Red Team Engineering notes, research references, and field notes on AI systems, responsible AI, governance, and compliance.
Topic

Adversarial ML

Adversarial machine learning attacks, taxonomies, and mitigations across the ML lifecycle.

adversarial mlevasionpoisoningprivacy attacknist ai 100-2
Evergreen Overview

Adversarial ML gives useful language for thinking about evasion, poisoning, privacy attacks, and model misuse. It becomes especially valuable when teams need to distinguish classic model risk from the broader application-layer risks found in modern LLM systems.

What this page helps separate
  • Model-centric attacks versus system-centric AI failures
  • Poisoning, evasion, privacy, and misuse in practical terms
  • Where adversarial ML taxonomies still help in generative AI discussions
Why it matters operationally
  • It creates cleaner language for policy, governance, and technical review
  • It helps teams choose the right controls for the right threat model
  • It prevents application-layer AI risk from being described too vaguely
Who this page is for
  • Security teams linking GenAI issues to broader model risk
  • Researchers working across predictive and generative AI
  • Practitioners who need disciplined taxonomy and terminology
References

Current notes, events, and source material

These items are included because they add useful evidence, framing, implementation detail, or upcoming context for teams working in this area.

IEEE ACSAC December 7, 2026 - December 11, 2026 event upcoming

IEEE ACSAC 2026

The Annual Computer Security Applications Conference brings applied security research to Los Angeles and includes four especially relevant workshops: agentic AI in offensive and defensive cyber operations, trustworthy machine learning, security and safety of AI in biotechnology, and AI for cyber-threat intelligence.

Agent SecurityAdversarial MLAI Red Teaming
View details Open event page
ACM AISec November 15, 2026 - November 19, 2026 event upcoming

ACM AISec 2026

The 19th ACM Workshop on Artificial Intelligence and Security is co-located with ACM CCS in The Hague and covers adversarial ML, model privacy, AI misuse, foundation-model red teaming, dangerous-capability evaluation, deceptive alignment, multi-agent risk, and autonomous-agent containment. Organizers have not yet assigned the workshop's exact day within the November 15–19 conference window.

Adversarial MLAI Red TeamingAgent Security
View details Open event page
USENIX Association August 12, 2026 - August 14, 2026 event upcoming

USENIX Security '26

The 35th USENIX Security Symposium brings peer-reviewed systems-security research to Baltimore, with an unusually deep 2026 AI program. Dedicated ML-security sessions cover real-world and indirect prompt injection, adaptive jailbreak attacks, agent memory and resource-abuse failures, LLM privacy and backdoors, and a systematization of agent attacks and defenses; proceedings are open access.

AI Red TeamingAdversarial MLAgent Security
View details Open event page
IEEE Security & Privacy May 18, 2026 - May 21, 2026 event event archive

IEEE Symposium on Security and Privacy 2026

The 47th IEEE Symposium on Security and Privacy and its workshops convened in San Francisco with a dedicated machine-learning security track and research on LLM privacy, alignment, jailbreaks, attacks, and defenses. The co-located Secure Agents for Generative AI workshop concentrated specifically on agent architectures and their security boundaries.

Adversarial MLAI Red TeamingAgent Security
View details Open event page
IEEE SaTML March 23, 2026 - March 25, 2026 event event archive

IEEE SaTML 2026

IEEE SaTML 2026 brought secure and trustworthy machine-learning research to the Technical University of Munich. The program addressed attacks and defenses, system and algorithm verification, ML privacy and forensics, interpretability and fairness, trustworthy data curation, and practical evaluation of learning systems.

Adversarial MLModel EvaluationAI Red Teaming
View details Open event page
Internet Society / NDSS February 23, 2026 - February 27, 2026 event event archive

NDSS Symposium 2026

NDSS 2026 combined systems-security research with dedicated AI-security sessions and the co-located LAST-X workshop on LLM-assisted security and trust exploration. The archived program includes attacks on LLM availability and alignment, watermark robustness, AI-enabled phishing defenses, adversarial ML, and evaluations of LLMs used in security work.

AI Red TeamingAdversarial MLAgent Security
View details Open event page