Paper 4: Behavioral Governance of Artificial Workers

ADI Research Series released Paper 4, "Behavioral Governance of Artificial Workers: Evaluating Completed Work Under Declared Authority and Behavioral Requirements," as a public preprint on Zenodo under DOI 10.5281/zenodo.23887999. The paper examines whether a completed AI worker output can be evaluated against declared authority, scope and behavioral requirements before release, with external controls deciding whether a proposed completion meets a stated acceptance contract. The author frames the finding as a bounded evaluation claim for the tested setting, not a claim of universal safety, deterministic language generation or general intelligence equivalence, and invites critique of the methods, acceptance boundary and limitations.

I have released Paper 4 in the ADI Research Series: Behavioral Governance of Artificial Workers: Evaluating Completed Work Under Declared Authority and Behavioral Requirements . The paper asks a narrower systems question than whether an AI worker can finish a task: can a completed output also be evaluated against declared authority, scope, and behavioral requirements before it is released? The model may generate a proposed completion. External controls decide whether that completion meets the stated acceptance contract. This is a bounded evaluation claim for the tested setting—not a claim of universal safety, deterministic language generation, or general intelligence equivalence. Public preprint and DOI: Behavioral Governance of Artificial Workers: Evaluating Completed Work Under Declared Authority and Behavioral Requirements | Zenodo https://doi.org/10.5281/zenodo.23887999 I welcome critique of the stated methods, acceptance boundary, and limitations.