Loading…
Concept
AI Egress Control
Egress control decides what may leave, before it leaves. The destination, the classification of the payload, and the actor's entitlement are all knowable at the call — which matters because exfiltration is usually the composition of two permitted capabilities rather than one forbidden act.
How do I stop an AI agent sending data where it should not?
Egress control decides what may leave, before it leaves. The destination, the classification of the payload, and the actor's entitlement are all knowable at the call — which matters because exfiltration is usually the composition of two permitted capabilities rather than one forbidden act.
2 patterns answer this
Control MCP egress
An agent calls a tool that reaches a network destination, uploads a file, or posts data to a third party.
8 executable policies
Third-party data sharing
A workflow posts customer or business data to a webhook, an external API, or a vendor integration.
6 executable policies
From this term to a governed action
- Control MCP egressThe recurring action behind this term.
- Copilot Agent Tool WhitelistAn executable policy that implements it.
- Try to break it in the sandboxA real signed verdict, no account.
- Deploy on your surfaceAvailable on 6 platforms today.
- Verify the proof yourselfCheck the signature without trusting us.
- Start in Shadow ModeMeasure on your traffic before enforcing.
Where this fits
Decionis is the deterministic decision layer between intent and execution. This concept is one way in; the architecture behind it is the same in every case.