Where sensitive data leaves, and what happens next.
Data walks out of an organization through a handful of paths: AI tools, email, cloud drives, the agents your developers run, and the laptops in between. Each one below says what it does for you, and Disclosures says how far each one is proven today.
Six things RedactWall does for you
Two decide before data moves, one reports on what is already there, and two run where no server sits in the path: on the laptop and inside your developers’ agents. The sixth is what makes the rest provable.
One set of rules, everywhere
No surface runs looser than the others.
Approved before it takes effect
Your risk owner signs off before anything enforces, and 21 hard stops cannot be turned off by anyone.
Built for regulated data
RedactWall finds identity documents, account and routing numbers, and payment data, and it matches against your own records. Coverage goes deepest for credit-union member and loan data.
See what that covers- Gateway-compatible applications keep their provider protocol after you configure the RedactWall route. Endpoint, browser and MCP paths require their named enrolled components.
- The audit record stores decision metadata, including the kinds of finding and the action taken, rather than prompt bodies. Employee activity can retain encrypted content under separate access and retention controls. Read the training policy and its deployment status.
- Coverage published surface by surface. Read the coverage table
Prove any of it to whoever asks
A record of decisions from the governed paths you deploy, and a way for your own team to detect entry changes and covered tail removal.
Start with a defined pilot
Confirm the proposed browser and gateway scope, verify the first device, then expand within the agreed cohort.
- Pilot starts with one governed path
- Each enabled surface reports to the same console
- Coverage and proof status are published