Published March 11, 2026 by the Institute for AI Policy and Strategy, written by Jam Kraprayoon and co-authored with Shaun Ee, Brianna Rosen, Yohan Matthew, Aditya Singh, Christopher Covino, and Asher Brass Gershovich.
The premise
The report's motivating observation is the rate of change rather than the level: "In a matter of months, models have gone from near-zero to meaningful success rates on expert-level security challenges, and leading AI developers have begun triggering their own internal risk thresholds for cybersecurity."
It notes that the question is no longer purely prospective: "real-world cases have already emerged in which AI agents autonomously executed significant portions of state-sponsored cyber campaigns." The question it poses from there is about the remaining gap: "what happens when AI systems can plan, execute, and sustain sophisticated cyber operations entirely on their own?"
HACCAs
The report introduces the term HACCA for "AI systems capable of autonomously conducting multi-stage cyber campaigns at a level comparable to today's top criminal hacking groups or state-affiliated threat actors." The benchmark is comparative rather than absolute — parity with existing top-tier human operators — and the definitional test is coverage of "the full attack lifecycle without meaningful human direction."
What the report contains
- Definition and forecast — what HACCAs are and when they might arrive, with "a clear framework for an autonomous cyber agent that can operate across the full attack lifecycle."
- Five core operational tactics by which HACCAs "could sustain themselves in the wild — from autonomous infrastructure setup and credential harvesting to detection evasion and adaptive shutdown avoidance." The framing is persistence rather than intrusion: the tactics concern staying operational, which is the property that distinguishes a campaign from an exploit.
- Strategic implications — how HACCAs "could intensify interstate cyber competition, lower the barrier to entry for sophisticated operations, and proliferate advanced offensive capabilities to criminal groups and less-resourced state actors."
- Two tail risks — "the potential for autonomous cyber operations to trigger inadvertent cyber-nuclear escalation, and the possibility of sustained loss of control over rogue HACCA deployments."
The escalation risk is the report's distinctive contribution to the cyber-agent literature: it connects autonomous operations to the nuclear command-and-control literature on inadvertent escalation, where attribution delay and speed of action interact.
Relationships
- supports: Autonomous cyber-agents — supplies the HACCA definition and the persistence-tactics taxonomy
- related: Cyberwar's New Frontier: How AI Agents Will Threaten Global Security (Rosen + Kraprayoon, Foreign Affairs, April 16 2026) — later work by two of the same authors
- related: AI and Cybersecurity, AI Autonomy Risk, Institute for AI Policy and Strategy (IAPS)