Skip to content
AIGP

RFC-020: Governed Autonomy, Symbolic Intent, and D-DNA Evidence — 8. Architectural Deterrence and Inhibition

AIGP Specification › RFC-020: Governed Autonomy, Symbolic Intent, and D-DNA Evidence › 8. Architectural Deterrence and Inhibition

← 7. Human Language Is Not a Control Plane · Section index · 9. Autonomous Governance Laws →

8. Architectural Deterrence and Inhibition

Machines are not deterred by punishment. Therefore, AIGP defines machine deterrence as architectural inhibition.

Inhibition may include:

deny action
degrade autonomy
disable tool
disable actuator
disable force capability
return to base
pause mission
isolate actor
revoke authority token
quarantine governance state
require human reauthorization

Example:

inhibition_controls:
  if signature_invalid:
    action: reject_command
    ddna_event: command.rejected.invalid_signature


  if authority_expired:
    action: degrade_to_safe_mode
    ddna_event: authority.expired.degraded


  if envelope_violation:
    action: stop_or_return
    ddna_event: envelope.violation.enforced


  if force_action_without_human_confirmation:
    action: disable_force_capability
    ddna_event: force.disabled.no_human_authority


  if repeated_governance_conflict:
    action: quarantine_actor
    ddna_event: actor.quarantined.governance_conflict

AIGP doctrine:

Policies encourage humans. Constraints inhibit machines.

← 7. Human Language Is Not a Control Plane · Section index · 9. Autonomous Governance Laws →