The HAOP Method: ARECC Applied to AI-Enabled Work Systems

The HAOP Method adapts ARECC — Anticipate, Recognize, Evaluate, Control, and Confirm, industrial hygiene's established decision-making framework — to the design of AI-enabled work systems. The sequence is inherited; the work-system translation is the contribution. The adaptation changes the unit of analysis from the AI product to the interaction of human, AI, and organizational performers across an operating workflow, and connects each phase to True Function, work-as-done, function allocation, consequential transitions, Verification Requirement and Verification Capacity, Verification Gates, grounding, protected pause authority, and continuing revalidation. It introduces the AI Deployment Safety Data Sheet (ADSDS), a bidirectional instrument serving as hazard communication to the adopting organization and as a requirements specification for the AI performer, within a workflow-level translation of the hazard communication elements. The result is a disciplined way to decide not only how to control an AI deployment, but whether a function should be delegated at all. This concept note is a short-form public deposit in the HAOP development series. Companion deposits: Human, AI, and Organizational Performance (HAOP): A Safety Framework for the AI Era, v5.0 (DOI: 10.5281/zenodo.21868036), the framework this method applies; From Human-in-the-Loop to Human-in-the-Design, v1.1 (DOI: 10.5281/zenodo.21633748), supplying the Verification Requirement and Capacity apparatus and the three design obligations; Verification Overrun: The Safety Hazard Inside AI Oversight (concept DOI: 10.5281/zenodo.20874288; resolves to the latest version); The Collapse of Correction: Closed Informational Loops in Safety-Critical Work, v1.0 (DOI: 10.5281/zenodo.20856122).