From the Local Government edition of September 24, 2026
Metro Nashville publishes public AI Use Registry accurate as of Sept. 9, 2026
Metropolitan Government of Nashville and Davidson County (Metro ITS) · County/metro transparency registry for municipal AI systems · Nashville/Davidson County, Tennessee, United States
- Publisher
- Nashville.gov (Metro ITS)
- Original publication
- Metro AI Use Registry page stating inventory accurate as of 9/9/2026
- Source retrieved
- 2026-09-25
- Event date
- 2026-09-09
What happened
Metro Nashville's public Artificial Intelligence (AI) Use Registry states the inventory is accurate as of 9/9/2026 and will grow as systems are added. Entries describe purpose, department, vendor/platform, AI technology type, decision role, human oversight, and status. Examples spanning local-government services include Adobe AI and Microsoft Copilot (enterprise active, advisory/human-in-loop), Granicus Government Experience Agent Meetings Agent pilot answering public meeting/legislation questions from Legistar with source links and staff monitoring, Archistar AI PreCheck planned/evaluation for Codes building-permit completeness and zoning/building-code pre-checks with humans retaining final authority, CYVL pavement/ROW asset intelligence planned/pilot, and DocAccess accessibility remediation planned/pilot. Many entries explicitly label Decision Role as Advisory (Human in Loop) and Human Oversight as Staff Review Required. Why now: a living U.S. metro inventory dated nine days before this edition date models the transparency control Fort Worth's audit says is hard—and that chatbot procurements (e.g., Grand Rapids RFP due 9/24) should emulate.
Why it matters
Concrete U.S. local-government public AI registry covering permitting pre-check, legislative chat, accessibility, and staff copilots with explicit human-in-loop labels.
Evidence and measured results
Nashville.gov registry: accurate as of 9/9/2026; fields for purpose/department/oversight/status; Archistar permit PreCheck planned with human final authority; Granicus GXA Meetings Agent pilot with source links; Copilot/Adobe active advisory; multiple Staff Review Required labels.
Limitations and uncertainty
Registry is Metro self-description; statuses include planned/pilot/active without measured outcome metrics; some public-safety tools appear in the inventory but are not the focus of this local-government edition's analysis; accuracy date is 9/9/2026, not a same-day update.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-09-25; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- AI tools proliferate across departments without a public, dated inventory councils can trust.
- Stakeholders
- CIO/CISO
- Codes/permitting
- Clerk/Council office
- 311/digital services
- Privacy/legal
- Discovery questions
- What is the registry as-of date?
- Are decision-role and oversight mandatory fields?
- Must new chatbots be listed before launch?
- Who verifies entry accuracy?
- Value hypothesis
- A public dated registry plus human-in-loop labels improves accountability for permitting and resident AI buys.
- Potential engagement
- Registry schema, first inventory load, and pre-launch listing gate for chatbot/permit AI—not a model vendor bake-off.
- Unsupported claims
- Registry proves tool accuracy
- Planned Archistar equals measured permit speedups
- Human-in-loop labels guarantee review quality
- All Metro AI is listed exhaustively
Pre-sales engineering
Role takeaway
- Fit
- Public AI system registry with oversight taxonomy across chat, CV, and copilots.
- Architecture and integration
- Department submits entry → ITS/security review → publish HTML/API → update on status change.
- Prerequisites
- Field taxonomy
- Department responders
- As-of dating
- Change process tied to procurement
- Constraints
- Self-attestation
- No outcome metrics in entries
- Mixed sensitivity of listed systems
- Update lag risk
- Security
- Avoid publishing sensitive model/data details; still require internal risk reviews before public summary.
- Proof of value
- Percent of AI projects listed pre-production
- Median update latency
- Field completeness
- Public as-of freshness
Delivery
Role takeaway
- Work
- Inventory campaign, editorial review of entries, publish with 9/9-style as-of stamp, and gate go-lives on listing.
- Dependencies
- Metro ITS
- Codes (Archistar)
- Clerk/Council (Granicus)
- Department AI owners
- Ownership
- ITS maintains the registry platform; departments own entry accuracy; security owns prohibited-use language.
- Skills and adoption
- Train program managers to fill decision-role/oversight fields honestly.
- Governance checkpoints
- Pre-launch listing
- Quarterly as-of refresh
- Spot audits of active entries
- Retirement of decommissioned systems
- Acceptance criteria
- As-of date displayed
- Human oversight field populated
- Permit/resident bots listed before production
- Change log or refresh cadence documented
- Risks
- Stale registry
- Rubber-stamp human-in-loop
- Omitting shadow tools
- Public confusion of planned vs active
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Registry is metadata about heterogeneous systems (RAG chat, computer vision, copilots)—not a single platform—so integration patterns stay use-case-specific.
Governance
Who approves, reviews and stays accountable for outcomes?
Publishing decision-role and oversight fields forces departments to declare whether AI is advisory; inventories must be dated and updated.
Security and privacy
What data, permissions and controls need testing?
Entries that ban scoring/surveillance uses (e.g., Meltwater limits) show registry text can encode prohibited purposes alongside allowed ones.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
DocAccess-style AI accessibility tooling appearing in the registry shows workforce and ADA remediation as first-class AI use cases.
Procurement
What should contracts, pricing and exit terms secure?
Require registry listing and oversight fields in chatbot/permit AI contracts before notice-to-proceed.
Operating model
Which teams own the service once it runs?
ITS-hosted registry with department-owned truth creates a clear RACI for sprawling AI portfolios.
Publication history
- 2026-09-24Local Government · Issue 194 resources
Stable resource ID: nashville-metro-ai-use-registry-20260909