Traefik Labs Launches Sovereign Trust Plane for Verifiable AI Agent Governance
At AGNTCon + MCPCon Europe, Traefik Labs announced the Sovereign Trust Plane (STP) — a new capability in Traefik Hub that provides cryptographically verifiable evidence of what AI agents were authorized to do, under which policies, and whether those records have been altered.
The Problem: Black Box Agents
As I discussed in my AGNTCon PreDay recap, one of the biggest challenges in productionizing AI agents is traceability. When an agent has write access to your CI/CD pipeline, your database, or your cloud infrastructure, you need more than logs — you need cryptographic proof of what happened and why.
Traditional observability tools capture what happened. The Sovereign Trust Plane aims to prove what was authorized to happen, what actually happened, and whether the record can be trusted.
How Sovereign Trust Plane Works
The Sovereign Trust Plane operates as a verifiable governance layer within Traefik Hub:
- Authorization Evidence — Before an agent takes any action, the STP records the policy decision that authorized it, signed with cryptographic proofs
- Action Audit Trail — Every tool call, API request, and state change made by the agent is logged with a tamper-evident signature
- Record Integrity — The STP uses a Merkle tree structure to ensure that any attempt to alter historical records is detectable
Sudeep Goswami, Chief Product Officer at Traefik Labs, explained the motivation: “Enterprises adopting agentic AI need to reconcile two competing demands: agents need autonomy to be useful, but enterprises need accountability to be safe. The Sovereign Trust Plane gives them both.”
Integration with MCP and Agent Gateways
The STP integrates directly with MCP endpoints and agent gateways, meaning it works with the open standards stack that’s gaining traction at events like AGNTCon. This is significant because it means teams can deploy verifiable governance without replacing their existing agent infrastructure — the STP sits as an orthogonal layer that observes and attests to agent behavior.
Availability
The Sovereign Trust Plane is available in Traefik Hub as of today. It supports MCP, OpenAI-compatible API, and standard agent frameworks including LangGraph and AgnTCY.
