Enforcement
Rights registry infrastructure. Records rights, preferences, declarations, or provenance for later lookup.
Practical force depends on reliable identity, accurate declarations, adoption, and downstream use of the registry.
Catalog status describes public availability, not legal validity, adoption, or proven effectiveness.
What it is
Spawning’s Do Not Train Registry is an opt-out system for creators who do not want their works included in future AI training datasets. In practice it has been closely associated with the Have I Been Trained workflow for discovering indexed works and recording training preferences.
The registry turns individual opt-outs into reusable signals, while the current public implementation appears to be in a transitional state.
Spawning Do Not Train Registry records rights, preferences, or provenance in a registry for images across the collection and training stages. It also incorporates preference signaling. Registry usefulness depends on accurate declarations, identity and rights verification, and adoption by the systems expected to consult it. Public materials describe an in-progress proposal or implementation; the newest dated source in this profile is “Spawning API documentation continues to describe the Do Not Train Registry” (July 16, 2026). These details describe the published mechanism and evidence, not a finding about legal validity, adoption, or effectiveness.
Limitations
The Have I Been Trained site is currently under maintenance, while Spawning's API documentation still describes access to the registry. The registry only works when model builders choose to honor it in dataset curation and training workflows. Earlier Spawning blog posts about the project are no longer available at their cited URLs.
Evidence trail
- Jul 16, 2026 · Primary sourceSpawning API documentation continues to describe the Do Not Train Registry
- Jul 09, 2026 · Primary sourceHave I Been Trained site reports that it is under maintenance
Adoption signals
No public adoption figure found.