Your agent reads a message, asks a model, and calls a payments API. Months later someone asks why. Trustvice 360 records each run in an ordered, tamper-evident chain that an auditor can check without access to your systems.
TrustActorNexusVectorNotarianTOPSLensApplication logs can say what happened. They cannot show that the log was not changed after the question was raised, because whoever controls the log store can rewrite it.
Which model, with which prompt, produced this decision?
What did the agent actually send to the payments API, and what came back?
Can you show these logs were not edited after the dispute started?
The first two questions are about content; the third is about trust. A tamper-evident chain gives each logged fact an order, a time and a signature that the owner of the log cannot quietly rewrite or backdate.
AppliesSince 25 May 2018
A person has the right not to be subject to a decision based solely on automated processing that produces legal or similarly significant effects, with exceptions. Where an exception applies, the controller must provide safeguards, at least the right to obtain human intervention, to express a point of view and to contest the decision.
Official text ↗AppliesDepends on the system’s risk class
If your agent is part of a high-risk AI system, the AI Act adds logging and log-retention duties with their own dates. They are set out on a separate page.
Read more →AppliesOngoing
Internal audit, enterprise customers and insurers increasingly ask the same three questions about any agent that can move money or data. Answering them with evidence another party can check is faster than answering them with screenshots.
Legal references are summarised and dated as of October 2026. They are given so you can see where the need comes from, not as legal advice. Read the official text, and ask counsel how it applies to you. Trustvice does not make any organisation compliant with any law.
Your code posts each agent run to a TrustActor actor, such as Witness, with one call to the Agentic API. The run carries the agent's DID and the run payload.
The actor observes the run, decides against the rules you declared and anchors the result in your workspace's hash chain: an ordered, tamper-evident record of what each actor observed and decided.
Notarian signs snapshots of the audit chain with Ed25519, with a Merkle root over the records. TOPS, the transparency chain, seals hourly and anchors the seal with OpenTimestamps.
An auditor can verify the chain from a starting sequence, replay a run from its chain entry, diff two runs and check snapshot anchors on the public verify path, without an account. Lens turns a query over the audit chain into a signed evidence bundle.
Each suite covers one part of the record. Open a suite to see what it does.
Actors such as Witness and Conscience check each run against declared rules and record it in the workspace hash chain, with replay and diff.
Open suite →Policy and trust engines for agent behaviour, such as Charter, Sentinel and Tribunal, to declare the rules runs are checked against.
Open suite →Edge engines in front of the agent's API traffic, such as Policy, Compliance and Routing.
Open suite →A capability of Trustvice 360: Ed25519-signed snapshots of the audit chain, anchored and publicly verifiable.
Open suite →A capability of Trustvice 360: the append-only transparency chain, sealed hourly and anchored with OpenTimestamps.
Open suite →A declarative query over the audit chain that produces a signed evidence bundle for an auditor.
Open suite →Harbor Lending's claims-triage agent refunds a late fee. The run is posted to the Witness actor and lands in the workspace hash chain; later, an auditor replays it from its chain entry.
The questioncurl -sS https://api.trustvice.com/v1/actions/trustactor.run \
-H "Authorization: Bearer $TRUSTVICE_TOKEN" \
-H "X-Workspace-Id: ws_…" \
-H "Content-Type: application/json" \
-d '{
"version": "v1",
"actor": "witness",
"agentDid": "did:web:harborlending.example:agent:claims-triage",
"input": { "task": "refund", "loanId": "LN-20418", "amountCents": 2500 }
}'
→ { "ok": true, "data": { "status": "ok", "receipt": { "runId": "run_…", … } } }{
"runId": "run_…"
}
→ { "ok": true, "data": { "ok": true, "found": true } }
trustactor.runs.verify-chain { "from": 1 }
trustactor.runs.diff { "a": "run_…", "b": "run_…" }Request fields as documented: /docs/api/trustactor
A signed record is evidence of a narrow fact: these values were recorded by Trustvice at this time, and have not changed since. It is not a certificate, an audit, a qualified timestamp or legal advice, and it does not by itself satisfy any legal obligation.
Every workspace can reach every suite; plans differ by capacity. Read the docs, or talk to us about what you need to prove.
Agents, evidence, governance, commerce and disclosure are all in Trustvice 360. Sign up free and open every suite from one dashboard.