TNQ Research is the standing analytical practice at TrueNorth Quantum. We publish a monthly catalog of reports, methodology references, field memos, and position papers — sourced from active engagements, drawn from the operating ledger of the Northern Shield platform, and intended for technical, regulatory, and executive audiences operating where AI agents now do consequential work.
Featured release · June 2026
POV-2026-002Perspective · Insurance & underwriting2026-06-0917 pp · PDF
Insurance-grade evidence: what underwriters actually want from an AI agent audit trail.
An audit trail built to satisfy your engineers is not the same as one built to satisfy an underwriter. The engineer asks whether the system worked. The underwriter asks something colder: when it fails, can the loss be attributed, bounded, and proven — and therefore priced. This perspective sets out the eight properties that make an AI audit trail insurance-grade, the regulatory and market context that has produced them — including the ISO January 2026 endorsements carving AI exposures out of general-liability cover — and why the underwriter's checklist turns out to be the governance checklist, written in the language of money.
Authored by TNQ ResearchClassification · Release 1.0 · Public
Document spine
Section 1 — The audit trail has a reader you have not met
Section 2 — What the underwriter is actually buying
Section 3 — The properties that make evidence insurance-grade
Section 4 — Why agents make the trail the only witness
Section 5 — How to think about it: write for the hostile reader
Section 6 — The market is already on this path
Section 7 — Build the testimony before you need it
A perspective on the eight properties that distinguish an audit log from insurance-grade evidence — attributability, tamper-evidence, completeness, contemporaneity, decision provenance, boundedness, reproducibility, and independent verifiability — and why the ISO January 2026 endorsements make this the operating standard whether you are buying cover or not.
An architectural reference for parallel-trained oversight in the regulated enterprise. Surveys the regulatory geography — EU AI Act, SEC enforcement, FDA, banking model-risk, NAIC — and the five-layer full-stack alternative to AI-alone deployment.
A complete enumeration of the engagement artifacts that constitute a TNQ Digital Employee build — phases 1 through 6, with sample structure for each artifact and acceptance criteria for the five sign-off milestones.
TNQ Research18 pp · PDF
MEM-2026-001
2026-05-08
Field memo · Engagement notes
Multi-track commission administration at infrastructure scale
Sanitised for release
Engagement notes from the deployment of a Digital Employee administering three parallel commission tracks across more than 100 active infrastructure projects, including a ten-year declining offtake stream with continuous reallocation.
TNQ Research12 pp · PDF
Coming soon
POV-2026-001
2026-04-29
Position · Industry direction
Why AI agents will need the same governance discipline enterprise IT eventually got
Release 1.0
A position paper arguing that the AI-agent industry is presently at the stage enterprise IT reached in the late 1990s — and that the governance and audit infrastructure now being retrofit at cost will be a prerequisite for any AI deployment of consequence within five years.
TNQ Research11 pp · Web · PDF
Coming soon
RPT-2026-002
2026-04-02
Report · Classification framework
The Risk Tier Framework: a five-dimension rubric for agent deployments
Release 1.0
A formal classification rubric for Digital Employee deployments across five risk dimensions — blast radius, reversibility, data sensitivity, regulatory exposure, external visibility — with calibrated control sets and oversight cadences for each of four resulting tiers.
TNQ Research22 pp · PDF
Coming soon
Distribution list registration
One release a month. No marketing.
Each monthly distribution contains one new piece of research published by TNQ Research — typically a report, methodology reference, or position paper — along with a short editorial note from the analyst.
No product announcements, no campaign sequencing, no third-party tracking.
You are added to a distribution list maintained directly by TNQ; you may unsubscribe with a single click.
Please enter a valid email address.
✓You’re registered. The next TNQ Research release will arrive in your inbox.
Registration data held by TrueNorth Quantum Inc. for the sole purpose of distributing TNQ Research releases. Not shared with third parties. Not used for marketing outreach. Governed by the TNQ privacy notice.
Request release
Insurance-grade evidence — full release.
Provide your details to download the complete perspective (PDF, 17 pp). Your information is used solely for distribution acknowledgement and TNQ Research follow-up.
Please complete all required fields.
Held by TrueNorth Quantum Inc. solely for TNQ Research distribution and follow-up. Not shared with third parties. Governed by the TNQ privacy notice.
✓
Your download has started.
If the download did not begin automatically, click below. A confirmation has been recorded.