State & Local GovernmentIssue 532 records
State Government
Two first-time sources from Oct 29-30. The U.S. House Oversight Committee asked Delaware for every record on its draft AI-run company sandbox, including internal AI chat logs and consultant agreements, as state lawmakers question its liability shield; the draft has not been introduced. In India, Jharkhand signed for an AI and analytics system linking about a dozen revenue portals to flag tax evasion, without published criteria for what triggers an investigation. Neither has measured outcomes.
- Evidence records
- 2
- Cross-source patterns
- 0
- Evidence classes
- 2 independent reporting
- Outcomes
- 1 cautionary1 emerging
- Source freshness
- 2 new this fortnight
- Research completed
- 2026-10-31
Choose a role to see its takeaway beside every record in the ledger.
Patterns across the evidence
Synthesis · Lighthouse Advisory interpretation
The evidence in this edition did not support a cross-source pattern. Each record below stands on its own.
Evidence ledger
Full record · every source keeps its link and limitations
U.S. House Oversight asks Delaware for all records, including internal AI chat logs, on its draft AI-run company sandbox as state lawmakers question liability shields and consultant ties
On Oct. 26, House Oversight Chair James Comer asked Gov. Matt Meyer for documents by Nov. 9 on Delaware's draft Artificial Intelligence Company (AIC), an entity run day to day by an AI agent whose capitalizer gets liability protection. The letter requests all communications, including internal AI chat logs, ethics and consulting agreements with outside contributors, safety-related records and all drafts, and says the Commission's two-page 2025 report did not mention assessing risks or consulting independent experts. Delaware Public Media reports the draft would run a 30-month sandbox with a state regulator and oversight committee; the state paid Nomos AI (Norm AI) $10,000 for consulting in September. Rep. Frank Burns said the framework will 'encourage riskier models to be used'; Rep. Krista Griffith said AIC actions would be logged and reviewed by an attorney and the AI Commission has not yet sent the draft to the governor or legislature. Meyer called the letter election-timed.
Why it matters, evidence and limitations
- Why it matters
- Why now: the federal request arrived this week with a Nov. 9 deadline, a substantive new development since our Sept. 24 Route Fifty item on accountability gaps in the draft bill (a separate archived source). It shows a state's AI policy process, its outside contributors and its staff AI chat logs becoming subject to federal records requests, a records exposure other states drafting AI law with industry input should plan for.
- Evidence and measured results
- Delaware Public Media article opened in full; the Oversight Committee letter PDF opened and read in full, including the four document requests. The draft 36-page bill and the Fortune op-ed cited in the letter were not reviewed. No evaluation of any AIC exists; this is policy scrutiny, not an outcome.
- Limitations and uncertainty
- The letter is partisan and timed days before an election, as the governor noted. Draft legislation has not been introduced and may change; the AI Commission still has questions about it. Consultant relationships are reported facts, not findings of wrongdoing. Delaware had not responded at publication.
Jharkhand's Commercial Taxes Department signs agreement for an AI and BI revenue intelligence system linking tax, mining, transport, land, excise and power portals to flag evasion
On Oct. 29, Jharkhand's Commercial Taxes Department signed an agreement, after a tender, with Choice Consultancy Services to build an Integrated Revenue Intelligence System (IRIS) under its CT-MIS project. IRIS will connect revenue-department portals, including GSTN APIs, the JIMMS mining portal, VAHAN transport, works accounts, land registration and property tax, labour cess, excise, trade licence and professional tax, and electricity billing, and apply business intelligence and AI tools to flag possible tax evasion and revenue leakage. Development is scheduled for six months followed by three years of implementation. The finance minister tied the project to an estimated Rs 22,000 crore shortfall he attributed to central policies. The report notes the release does not specify data fields, architecture or the criteria that will trigger an investigation.
Why it matters, evidence and limitations
- Why it matters
- Why now: signed Oct. 29 and not in our archive. It is a concrete state-level example of AI applied to cross-agency revenue data, a use case U.S. state revenue departments also pursue. Transferability is limited: India's GST network and portal landscape differ from U.S. state tax systems, and no results exist yet.
- Evidence and measured results
- News report opened in full; it relays a state government press release, which was not opened directly. Includes the reporter's own noted gaps on data fields, architecture and investigation criteria. No performance data.
- Limitations and uncertainty
- Single news report based on a government release; no technical specification, contract value or evaluation plan published. Revenue loss estimates are the minister's political claims without supporting calculations. Impact depends on implementation and departmental follow-up.
How to read this edition
Source findings, measured results and limitations come from the cited publications. Patterns, operating questions, role takeaways and implementation considerations are Lighthouse Advisory interpretation, stated as questions to validate locally rather than guaranteed outcomes. Vendor and operator claims are labeled as claims. Full research method.
- Independent reporting
- Independent reporting with attributable sources but without a formal evaluation design.