TaskQ turns every policy-bound AI request into exportable evidence: what ran, where it ran, which credential was used, which worker handled it, what was blocked, and whether the chain replays cleanly.
TaskQ should never pretend that software alone makes a customer compliant. The product provides enforceable controls and auditor-ready evidence for policy-bound AI execution.
Hard policy constraints are evaluated before a request reaches a worker. If no compliant capacity exists, TaskQ refuses with proof.
Placement decisions, credential access, blocked attempts, worker leases, usage, artifacts, and settlement all map into HashIQ entries.
Chain segments, placement proofs, registry snapshots, anchor proofs, and report packages are exportable by design.
The compliance plane turns customer defaults and request-level policy into release gates. Requests may tighten constraints. They may not loosen defaults.
Require execution in Canada, the EU, a customer-owned domain, or another approved jurisdiction.
Explicitly reject workers in disallowed jurisdictions even when they are faster or cheaper.
Restrict credential use to pinned, enclave, or customer-owned nodes.
Require evidence, audited, or TEE worker tiers for sensitive workloads.
A serious compliance report is not a pretty chart. It has to show policies enforced, workers used, credentials accessed, chain status, exceptions, anchors, and settlement reconciliation.
TaskQ records effective worker attributes at the time of placement. An auditor needs the state at execution time, not today’s dashboard state.
| Surface | What the report shows |
|---|---|
| Worker ref | Blinded or disclosed based on report scope and customer permissions. |
| Jurisdiction | Effective execution location, residency class, and excluded regions. |
| Attestation | Tier, evidence reference, expiry status, and weakest-link calculation. |
| Snapshot | Registry snapshot proof used for placement replay. |
The compliance surface needs to feel operational: lookup, report, export, verify, and inspect credential access.
GET /v1/chain/{request_id} GET /v1/chain/export?from=2026-06-01&to=2026-06-30 GET /v1/reports/audit?environment=prod GET /v1/credentials/{credential_id}/access # report outcome { "chain_verification": "passed", "anchor_verification": "passed", "placement_replay": "passed", "blocked_attempts": 17, "settlement_reconciled": true }
Policy enforcement. Credential access logs. Placement proof. Chain anchors. Exportable reports.