Scope: what this does not show
A design position that several parties converge on. It trades away capability and is not a measurement.
Corroborated: Supported by at least two independent sources.
Evidence
Jun 10, 2025
Jun 16, 2025
Oct 31, 2025
Jan 28, 2026
Feb 13, 2026
Oct 19, 2026
Microsoft extends Project Perception to treat AI agents as insiders, with agent identities, task-scoped permissions and risk-based access
Microsoft announces Project Perception features built on keeping controls outside the model, with task-scoped permissions and risk-based narrowing of access; a product announcement with no measurements.
How it relates to other findings
ReportedCorroboratedQualifiedContestedSupersededRevalidate· node size = evidence records · columns group by topic
Select a finding to see how it relates to others. Arrows point from the newer finding to the one it supports, qualifies, contests, or supersedes.
- This finding supports Separating an agent’s control flow from untrusted data can give provable protection against specified forms of hijacking under explicit assumptions.CaMeL is one implementation of this principle.
Key questions that rely on this finding
- 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.
Status history
- 2025-06-10ReportedDesign patterns paper argues for constraining agents. · record
- 2025-06-16CorroboratedIndependent framing of the same principle as the lethal trifecta. · record
- 2026-09-25CorroboratedcorrectionWillison's post quotes and builds on the design patterns paper, so it is not independent of it. Corroboration rests on separate organizations adopting the position, such as OpenAI's deterministic Lockdown Mode. · record