Enforcement
Voluntary preference signal. Communicates a requested condition or restriction to automated systems.
It only affects actors that detect and honor the signal; it does not itself prevent reuse.
Catalog status describes public availability, not legal validity, adoption, or proven effectiveness.
What it is
trust.txt is a machine-readable file used by publishers to declare affiliations and other trust metadata. In April 2024, the specification added a datatrainingallowed= variable, extending the framework into AI-related permissions as well.
The AI-use field adds training-permission metadata to an existing publisher coordination system rather than requiring a separate AI-only standard.
trust.txt publishes or proposes a machine-readable preference signal for web content at the training stage. It also incorporates protocol standardization. The signal communicates requested conditions; compliance depends on discovery, interpretation, and voluntary support by downstream systems. Public materials describe a currently available initiative; the newest dated source in this profile is “JournalList continues to publish the trust.txt specification” (July 16, 2026). These details describe the published mechanism and evidence, not a finding about legal validity, adoption, or effectiveness.
Limitations
trust.txt is publisher- and journalism-centric infrastructure; the AI-related signal is only one part of a broader trust-and-affiliation framework.
Evidence trail
- Jul 16, 2026 · Primary sourceJournalList continues to publish the trust.txt specification
- Feb 21, 2025 · Primary sourceBrowser extension launch described the network at about 3,000 publishers
- Apr 04, 2024 · Primary sourcetrust.txt spec added the datatrainingallowed field
Adoption signals
Users
about 3,000 participating publishers