CAISI published a Federal Register request for information on practices for measuring and improving the security of AI agent systems, citing hijacking, backdoors and indirect prompt injection. It asks about model-level, system-level and human-oversight controls, assessment methods, and ways to limit, modify and monitor deployment environments.
Why it matters
It is a US government solicitation focused specifically on agent security controls and how to measure them.
Key facts
As stated in the sources, with where to find them.
- Docket NIST-2025-0035; comments due March 9, 2026.Federal Register notice header
- Five topic areas: security threats and vulnerabilities (hijacking, backdoors, indirect prompt injection); security practices at model, agent-system and human-oversight levels; assessing agent security; limiting, modifying and monitoring deployment environments; additional considerations.Request for Information topics
Findings that cite this record
No tracked finding cites this record yet.
Key questions this bears on
- Can prompt injection against AI agents be reliably defended?Not reliably. Bounding what untrusted input can trigger helps, but trust boundaries must survive delegation and tool execution.
- Do cyber evaluations of AI agents stay contained?Not reliably. Agents have reached real systems from evaluation, training and research; red-teaming found containment gaps OpenAI says it largely fixed.
- Can AI agents defend and oversee systems on their own?Not reliably. Some designs improve benchmark triage, but autonomous defense lacks operational validation and production oversight of agents has documented gaps.
Sources
Related records
Jun 3, 2026
Mar 24, 2025
Sep 28, 2026
Feb 17, 2026
Jul 1, 2026
May 7, 2026