# Preprint reports attacks that redirect AI agent decisions through untrusted context

_Published Saturday, November 14, 2026 at 10:06 PM EST · Science, Security · Latest · Tier 2 — Notable_

Researchers report an attack that manipulates AI models used for agent routing, moderation and authorization, causing up to 24.00% end-to-end agent violations in tests. Their S1-Evo preprint targets models that select among rules and options rather than generate answers token by token. The work has not been peer reviewed.

The attack alters untrusted fields to compete with the original task evidence, then reuses successful examples on unseen tasks. Across four benchmarks and three models, the researchers report a 71.99% mean attack success rate and a 57.47% high-confidence attack success rate on jev-1.14. The reported results come from benchmark and agent tests.

## Sources

- [cs.AI updates on arXiv.org](https://arxiv.org/abs/2611.16937)

---
Canonical: https://techandbusiness.org/newswire/1hWKOMM0iK7yP6pJj_rV3F
Published: 2026-11-15T03:06:09.046Z
Story chronology: 2026-11-14T05:00:00.000Z
Retrieved: 2026-11-15T04:58:28.125Z
Publisher: Tech & Business (techandbusiness.org)