AI data governance
Document which data an AI workflow uses, who has permission to access it, and what happens when it moves between systems.
View AI governanceStart with the data flow
Identify the source of each input, the tools and services that receive it, and the outputs and logs that retain it. Review the access rules at each step.
Requirements to document
- Sources and quality. Where the information comes from and how errors or missing records are handled.
- Access and purpose. Which people, tools, and services are authorized to use each field, and for what task.
- Copies and retention. Where data is stored, logged, or shared, and how long it is kept.
- Review records. What evidence is needed to trace an action to its inputs and permissions.
Protocol research
Our TN-Proto work explores field-level encryption, signed records, and policy information attached to events. The implementation documentation describes how those mechanisms work.
Read more about AI governance and agent security.