4. Three Fundamental Questions¶

Note

[Author: introduce the three fundamental questions that structure remote attestation for AI agents.]

4.1. What to Attest?¶

[Author: discuss what should be attested for an AI agent — identity, code, configuration, model, runtime state, behaviour, and so on.]

[Author: discuss the mechanisms and protocols for attesting an AI agent.]

[Author: discuss the timing and triggers for attestation — at load, at runtime, per interaction, continuously.]

[Author: summarise the open research gaps across the three questions.]

[Author: summarise the standardization gaps across the three questions.]