Skip to content
AIGP

RFC-020: Governed Autonomy, Symbolic Intent, and D-DNA Evidence — 3. Problem Statement

AIGP Specification › RFC-020: Governed Autonomy, Symbolic Intent, and D-DNA Evidence › 3. Problem Statement

← 2. Motivation · Section index · 4. Design Thesis →

3. Problem Statement

Autonomous and agentic systems operate in dynamic environments where context changes continuously.

A system may transition within seconds from:

authorized → unauthorized
safe → unsafe
inside envelope → outside envelope
human present → human absent
comms healthy → comms degraded
confidence sufficient → confidence insufficient
mission valid → mission superseded
force prohibited → force explicitly constrained

Static policy is not enough for this environment.

The problem is not only whether a system “followed policy.” The deeper question is:

Was the autonomous action legitimate at that moment, under that authority, inside that envelope, according to that constraint, given that context, with replayable evidence?

AIGP treats this as a protocol problem.

← 2. Motivation · Section index · 4. Design Thesis →