Example ledger. Each decision is hashed, chained to the one before it, and signed, so an edit to any past record breaks the chain and is rejected.
Every agent decision, sealed as evidence
The Evidence Center turns each allow, block, and human approval into a tamper-evident record, maps it to the controls your auditors check, and puts status, certifications, and reports in one portal your reviewers can use on their own.
trust.agenttrustcore.com/enterpriseWhat's inside the Evidence Center
Everything procurement, security and legal teams ask for during an enterprise review — self-serve and always up to date.
Compliance & Certifications
SOC 2 Type II, ISO 27001, GDPR and HIPAA posture in one continuously-updated portal. Download reports on demand.
Audit-Ready Evidence
Every agent decision, policy check and human approval is logged to an immutable, exportable audit trail.
Data Protection
Encryption in transit and at rest, regional data residency, and privacy-by-design controls for sensitive workloads.
Real-Time Transparency
Live status, sub-processor lists, and security posture so your risk and procurement teams never wait on email.
Access Governance
Zero-trust access, SSO/SCIM, and granular role-based permissions across every agent and environment.
Deployment Assurance
VPC, private cloud, and on-prem deployment options with signed DPAs and enterprise SLAs.
From agent action to verifiable evidence
Every agent decision runs through the same trust pipeline — enforced, logged, and published so reviewers never have to take your word for it.
Agent acts
An AI agent requests a tool call, API action, or data access.
Policy check
The request is evaluated against your policies, roles, and guardrails in real time.
Log & sign
The decision, inputs, and outcome are written to an immutable, signed audit trail.
Publish to Evidence Center
Posture, evidence, and status surface in the portal for reviewers to self-serve.
Watch it allow & block in real time
Each example is a real agent action flowing through the policy engine. Switch agents to see how safe requests pass and risky ones get stopped before they reach a resource.
agent: Support Agent
action: issue_refund
target: Payments API
$8,400 refund to customer #4021
Evaluating...
Running policy checks against the requested action.
How the Evidence Center fits together
A layered system that sits between your agents and the outside world — enforcing policy, capturing evidence, and exposing it in one reviewer-ready portal.
Agent & App Layer
Your AI agents and applications connect via SDK, gateway, or proxy.
Policy & Enforcement Engine
Every action is intercepted, scored, and allowed, blocked, or escalated.
Evidence & Audit Store
Tamper-evident logs and compliance evidence, retained and exportable.
Evidence Center Portal
The reviewer-facing surface for status, certifications, and documents.
Certifications & frameworks
Our program is being independently audited. Reports and mappings will be downloadable directly from the Evidence Center at launch.
Questions
Evidence Center, answered
How can we prove to auditors what our AI agents did?
Evidence Center publishes signed records of every TrustCore decision and Veridict test result, alongside certifications, in a portal auditors and buyers can review directly.
Can Evidence Center replace security questionnaires?
It lets reviewers self-serve most questionnaire answers, certifications, and evidence, which shortens or removes many questionnaires.