Productised AI-governance services

How to engage

Every service below is available on inquiry — email hello@policywindow.org with the service ID, your organisation, the question you want answered, and any constraints (timeline, budget envelope, jurisdictions in scope). We respond within 5 business days with a fixed-scope proposal or an honest "not yet" if the engagement is outside current capacity. No pricing is published until a service is committed-to with a first customer (per charter §4).

Layer 2 — Paid professional

• Premium AI Governance Data Exports

  Available on inquiry

  Download structured, source-cited AI-governance data with stable IDs, confidence tiers, citations, snapshots, and machine-readable fields. Extends the free /wiki/catalog/json + /wiki/catalog/csv exports with provision-level anchors, source-defensibility status, and historical-snapshot bundles.

  For: Analysts · Researchers · Consultants · Journalists · Developers

  Free public exports already live at /wiki/catalog/json and /wiki/catalog/csv; premium tier adds provision-level fields + historical snapshots + SLA.

  Inquire about Premium AI Governance Data Exports →

Layer 3 — API + data

• AI Governance API

  Available on inquiry

  Programmatic access to Policy Window's structured AI-governance catalog: instruments, topics, concepts, coverage cells, enforcement cases, benchmarks, citations, confidence tiers, and historical states. Authenticated via scoped bearer tokens; commercial licence available.

  For: AI-governance tools · Compliance vendors · AI labs · Internal governance agents · Policy consultancies

  OpenAPI 3.1 spec live at /api/openapi documents the public catalog endpoints. Premium scopes (api.premium, change-feed.subscribe) gated on first commercial pilot.

  Inquire about AI Governance API →

• Regulatory Change Feed

  Scoped pilot

  Source-monitored change alerts and semantic diffs for high-priority AI-governance instruments. Tracks per-field changes to catalog rows + per-cell coverage updates; deliverable as RSS, JSON pull, or webhook push.

  For: Compliance teams · Policy teams · Consultancies · Journalists · AI-governance researchers

  RSS changelog feed live at /wiki/changelog/feed; semantic-diff infrastructure scaffolded; webhook delivery on roadmap.

  Inquire about Regulatory Change Feed →

Layer 4 — Procurement + compliance

• Procurement Evidence Packs

  Available on inquiry

  Source-cited evidence packs for AI procurement and vendor assessment, including relevant governance instruments, obligations, standards, benchmarks, enforcement cases, and RFP questions. Includes AI Act / NIST / ISO / OECD crosswalks and enforcement-risk summaries.

  For: Public-sector procurement · Foundations · NGOs · AI assurance teams · Policy consultancies

  Self-serve procurement checklist live at /wiki/for-procurement. Bespoke evidence packs require a 2-3 week turnaround per request; not legal advice.

  Inquire about Procurement Evidence Packs →

Layer 5 — Human-validated research

• Human-Validated Evidence Packs

  Available on inquiry

  AGI Social Scientist produces an evidence map + draft analysis; trained researchers validate key claims, context, methods, and interpretation before delivery. Output is a structured evidence object that writes back into the reusable evidence graph, not a one-off PDF.

  For: Foundations · AI-governance organisations · Think tanks · Policy labs · Policy consultancies

  First customer-gated. The expert validation bench is recruited per engagement; lead time is 4-6 weeks from commission. See /wiki/charter §4.5 for sponsor-firewall rules.

  Inquire about Human-Validated Evidence Packs →

• Expert Interview Add-On

  Available on inquiry

  A trained social-science researcher conducts a small number of expert or stakeholder interviews to validate, challenge, or contextualise an AGI-generated evidence map. Consent-gated, anonymisation-by-default, written protocol, declared data retention. No covert observational research; no human-subject work without an approved protocol.

  For: Evidence-pack customers extending findings with stakeholder voice · Policy labs validating regulatory transfer · Foundations evaluating programme impact

  First customer-gated. IRB-equivalent protocol template on roadmap; live recruitment of validators gated on first commission.

  Inquire about Expert Interview Add-On →

• AI Governance Implementation Case Study

  Available on inquiry

  A structured case study of how an AI-governance rule, framework, procurement requirement, or institutional practice is implemented in the real world. Outputs include case timeline, named actors, instruments invoked, interview evidence where applicable, implementation barriers, uncertainty, and a reusable structured case record.

  For: Foundations comparing implementation outcomes · Policy labs · AI assurance vendors · Public-sector procurement evaluating prior deployments

  Existing ImplementationCase + ResearchProvenanceRecord models scaffold the structured-output side; UI for public case-study pages is on the Batch B roadmap.

  Inquire about AI Governance Implementation Case Study →

• AGI Social Scientist Research Sprint

  Available on inquiry

  A fixed-scope, AI-native research sprint combining AGI-generated evidence mapping, source analysis, expert validation, and structured outputs. Productised — fixed scope, fixed deliverables, no open-ended consulting. Every sprint output writes back into the reusable evidence graph.

  For: AI-governance organisations needing rapid evidence on a defined question · Foundations scoping a new programme area · Policy consultancies augmenting existing client deliverables

  First customer-gated. Typical sprint: 3-4 weeks from commission to validated output. Sponsor firewall per /wiki/charter §4.5; no per-sponsor conclusions.

  Inquire about AGI Social Scientist Research Sprint →

What we do not offer

• Legal advice. Outputs are evidence references; consult qualified counsel for jurisdictional advice.
• Lobbying automation. No bulk regulatory comments, no synthetic public-consultation submissions, no astroturf campaigns (charter §7.5).
• Covert persuasion or political microtargeting. All editorial output is signed and dated (charter §7.6).
• Unapproved human-subject research. Expert interviews require informed consent, anonymisation, and a written protocol (charter §7.7).
• Open-ended consulting. Every engagement has fixed scope and structured deliverables that improve the reusable evidence graph.