CLAIM-2026-08-07-03

AGRI, a probe-gated reasoning defence that prepends anti-injection reasoning on demand, reduces attack success rate from 34.6% to 0% on Qwen3.5-27B on difficult AgentDojo settings while largely maintaining clean-task utility.

Why it matters

It shows the gap is closable with an ordinary engineering move — gate an extra reasoning pass on a detector — rather than a new model. That makes it a reasonable thing for a regulator or an auditor to expect of a deployer.

Sources

• arxiv2026your