Theory Audit — Field Note 07: Synthetic Conduct

Phase 2 — Theory Audit

Audit question

Is Synthetic Conduct a genuinely new governance problem requiring new theory — or is it a familiar configuration of agency, automation, language-act and responsibility concepts applied to AI systems? If a residual remains, what specifically is novel about the governance gap?

41 papers across 6 folders

Folder A: Artificial agency (6 papers). Folder B: Automation and human–machine systems (7 papers). Folder C: Language as action (6 papers). Folder D: Responsibility, causation and moral luck (8 papers). Folder E: AI security, adversarial prompting and emerging harms (8 papers). Folder F: Organisational authority and AI governance (6 papers).

Verdict: D — Governance construct retained provisionally

Synthetic Conduct is not a new mechanism. The residual is a governance gap: existing frameworks assign responsibility by tracing intent, authority or physical causation to a human or legal person. When an AI system produces consequential conduct through language — without a human directing each instance — that trace breaks down. The concept earns provisional retention as a diagnostic label for this gap, not as a settled causal mechanism.