PROVENANCE FIRST: Trust, Accountability, and Human Judgment in the Age of AI

Artificial intelligence is changing not only how work is produced, but how we decide whether that work deserves our trust. In Provenance First: Trust, Accountability, and Human Judgment in the Age of AI , Rodger P. Hildreth argues that the central challenge of generative AI is not simply detecting whether a machine was used. The more important questions are: What did the human contribute? What did the machine contribute? What was verified? What judgment was exercised? And who stands behind the result?

The book proposes a provenance-first approach to answering those questions. Rather than relying on AI detectors, vague disclosure statements, or impossible estimates of “how much” work was human or machine generated, provenance focuses on the creation process itself.

Hildreth develops Origin System , a practical framework built around three functions:

CAM — establishes the rules governing human–AI collaboration

TRACE — preserves meaningful evidence of how the work was created

ATTEST — identifies who reviews the work and accepts responsibility for it

At the center of the framework is the Origin Record , a selective account of material human and machine contributions, verification, corrections, approvals, and responsibility.

The argument extends across education, scientific research, law, healthcare, government, and creative work. Throughout, Hildreth emphasizes an equally important principle: provenance must not become surveillance. The goal is not to record every prompt, draft, or private thought, but to preserve enough evidence when the consequences justify knowing.

Provenance First does not claim that provenance proves truth, quality, originality, or competence. Its claim is more disciplined: better evidence about how consequential work was created can support better human judgment and clearer accountability.

As AI becomes embedded in ordinary intellectual work, the question will increasingly be not whether AI was used, but whether we know enough about the process—and the people responsible—to decide whether the result deserves our trust.

This version is grounded in the manuscript’s central argument, terminology, and applications.