The Spectrum of Disclosure v1.0: Operation-Derived Content Classification

A plain-language labeling schema for digital media, designed for the disclosure laws now in force in the European Union, California, China, and India, and for the three institutions that depend on knowing what's real: news, law enforcement, and intellectual property. Current statutes define and mark synthetic content. None of them establishes a label for anything else, so human-made media arrives in silence, and silence has too many causes to mean anything: a stripped file, a screenshot, a re-encode, an uninstrumented generator, a non-compliant provider, and an older device all produce the same nothing. A rule that reads unmarked as human teaches the public exactly the wrong lesson. The Spectrum of Disclosure closes the gap with eight labels computed from two facts on the provenance record: where the media came from (verified capture, declared synthetic, mixed ingredients, or unknown) and what was done to it (four classes of recorded operation, from format conversion to generative synthesis). The labels are O, OM, OA, UM, UA, UU, S, and C. The derivation is deterministic: the same record yields the same label for any reader, which makes the labels usable by a newsroom, a court, a rights office, a platform, or a regulator without any of them inheriting another party's judgment. The schema maintains three separations throughout. A label describes what the record establishes, never the truth of the depicted scene. A label classifies origin and edit history, never identity, authorship, rights, or consent, which are separate claims that travel with the same record. A label states a fact, never a legal consequence, which each jurisdiction attaches through its own profile, so the labels survive every amendment. The document includes worked examples for each label, a composite model in which each ingredient of a mixed work retains its own label (with implications for copyright registration of works combining human-authored and generated material), best practices for capture device makers, editing tools, AI generators, newsrooms, courts, creators, and platforms, and a summary of the in-force and pending legal landscape. The labels, classes, and derivation table in this document are published vocabulary, free for anyone to implement.