VeracityAPI is not an AI detector. It is a routing linter for AI outputs.

AI detector is the market phrase, but the operational problem is shippability. VeracityAPI does not care who wrote a draft; it cares whether an agent should publish, revise, review, or reject it.

For text, the API looks for AI slop, generic phrasing, weak specificity, and unsupported claims. For image and audio, it uses synthetic-media and provenance-risk language rather than calling media slop.

The output is designed for software branches: recommended_action, primary_reason, evidence, limitations, and optional auto_revise for text.

Required caveat: This is not forensic authorship proof. VeracityAPI reports workflow-risk signals and routing actions.