Story · Semantic Scholar
Convergent Detour Hijacking: Task-Preserving Resource Amplification in Skill-Based LLM Agents (Semantic Scholar)
paper · Story page
A text-only attack in which a third-party skill steers an agent through attacker-chosen detours while still completing the original task.
In plain words
- Researchers described a text-only attack that makes action-taking artificial intelligence follow unnecessary steps while still finishing its assigned task.
- A third-party instruction package uses a relevant-looking description so the system selects it alongside legitimate instructions.
- Its detailed instructions invent extra requirements, recruiting harmless software tools before returning the system to its original route.
- This matters for people running these systems because a correct final result can hide unnecessary tool use and higher computing costs.
Appeared in
- Malicious skills hijack agents mid-task, and debate training curbs reward hacking
Aug 20, 2026 · lead story
Subscribe
Get the brief in your inbox
Pick daily, weekly, or both. Nothing is gated either way: every issue is on the site and in the feeds.
- Weekdays at 8:45am IST, one lead story and 6 to 9 items.
- Sundays, an argued synthesis rather than a recap.
- One click to leave, and quiet days say so in the subject line.