State & Local GovernmentIssue 593 records
State Government
Maryland is rewriting how agencies buy technology, and Delaware's efficiency report quietly advances AI company legislation. Maryland DoIT launched a technical acquisition team and a problem-first intake door, and plans to use AI in acquisition only under the state's Responsible AI Policy while seven agencies test the process. Delaware's 2026 GEAR report says the AI Commission combined Artificial Intelligence Company and sandbox drafts into one bill for the General Assembly, with AICs denied voting rights, and notes staff use of an ARIS AI Companion plus a benefits-site chatbot. Tamil Nadu signed an MoU to stand up a PMU that will judge whether its State Data Centre and applications are AI-ready before migrating to a new facility. House Oversight's Delaware document request and North Carolina's GenAI RFP question deadline remain open.
- Evidence records
- 3
- Cross-source patterns
- 1
- Evidence classes
- 2 independent reporting1 government evaluation
- Outcomes
- 3 emerging
- Source freshness
- 3 new this fortnight
- Research completed
- 2026-11-06
Choose a role to see its takeaway beside every record in the ledger.
Patterns across the evidence
Synthesis · Lighthouse Advisory interpretation
States are hardening intake and commission rails before scaling AI buys or sandboxes
Maryland is forcing problem-first technology intake through a new technical acquisition team and will only introduce AI into acquisition under its Responsible AI Policy, while Delaware's GEAR report shows the AI Commission inventorying GenAI use and advancing a combined AIC/sandbox bill to the legislature. Both treat institutional rails—discovery, inventory, policy—as current work, with measured AI outcomes still ahead.
Operating questionBefore the next GenAI platform or agentic sandbox decision, does the state have problem-first intake, a GenAI inventory, and written policy gates that can stop or reshape the buy?
Supporting evidenceMaryland Department of Information Technology (DoIT)Delaware Government Efficiency and Accountability Review (GEAR); Delaware Artificial Intelligence Commission
Evidence ledger
Full record · every source keeps its link and limitations
Maryland DoIT launches a technical acquisition team and plans AI-assisted procurement intake under the Responsible AI Policy
On 5 November 2026 Maryland DoIT announced the state's first technical acquisition team and a strategic IT acquisition framework tied to the Procurement Reform Act of 2025. Agencies are to bring technology problems through a centralized "single front door" rather than naming a product first, so DoIT can test whether existing statewide systems already solve the need. FY2026 discovery found more than 1,700 technology requests representing over half a billion dollars in estimated demand, and only about 3% of IT buy-type and procurement-vehicle combinations have a purpose-built template, with some RFPs stretching to 118 pages. Nine DoIT review functions (including AI) are aligning standing compliance requirements. DoIT plans to partner with Partners for Public Good to use AI to streamline acquisition, subject to Maryland's Responsible AI Policy. The new process is in end-to-end user testing with seven agencies while DoIT recruits product managers, UX researchers, designers and engineers alongside procurement staff. No measured time or cost savings from the AI component are reported yet.
Why it matters, evidence and limitations
- Why it matters
- Direct U.S. state shared-services procurement reform with an explicit AI workstream under a Responsible AI Policy. Transferable to other states redesigning IT intake before GenAI platform buys; keep citizen 311/digital-service bots in Citizen Engagement unless the buy is enterprise acquisition tooling.
- Evidence and measured results
- DoIT press release dated 11/5/2026 states the technical acquisition team, single-front-door intake, FY2026 discovery figures (>1,700 requests; >$500M estimated demand; ~3% templated combinations; RFPs up to 118 pages), partnership plan with Partners for Public Good for AI in acquisition under the Responsible AI Policy, and seven-agency end-to-end user testing. StateScoop same-day reporting quotes CIO Katie Savage on problem-first discovery and corroborates the AI partnership and testing.
- Limitations and uncertainty
- AI for procurement is planned, not deployed with results. Discovery demand figures are estimates. Template and RFP-length statistics describe process pain, not proven savings from the new team. Recruiting is ongoing.
Delaware GEAR 2026 report: AI Commission combined AIC/sandbox bill for the General Assembly, GenAI inventory mandate, and ARIS AI Companion use
Governor Meyer's office released the 2026 GEAR Annual Report on 5 November 2026. Beyond the headline ~$117 million in estimated lifetime savings across 164 projects (planning estimates, not budget-booked results), the report's AI section states that House Bill 333 created the Delaware AI Commission with a mandate to inventory generative AI use across executive, legislative and judicial agencies and flag high-risk areas. Under House Joint Resolution 7, the Commission is working with the Secretary of State on an agentic-AI regulatory sandbox. Commission members reviewed separate draft bills on Artificial Intelligence Companies (AICs) and sandbox updates, combined them into one bill, and recommended presenting it to the General Assembly before the end of the legislative session; a major change explicitly denies AICs the right to vote. Operationally, the OneDelaware ERP team is using the ARIS Artificial Intelligence Companion to draft process-model frameworks, and the State Benefits Office launched "Benny," a website-navigation chatbot, in February 2026. The report does not say Delaware has answered the U.S. House Oversight document request due ~9 November.
Why it matters, evidence and limitations
- Why it matters
- Primary-source update on Delaware's state AI governance stack and the AIC/sandbox legislation that House Oversight has been probing. Relevant to state CIO, Secretary of State and continuous-improvement offices elsewhere; not a substitute for the still-pending oversight response.
- Evidence and measured results
- Opened 2026 GEAR Annual Report PDF (fetched 2026-11-05). AI Commission duties, HJR 7 sandbox direction, combined AIC/sandbox bill recommendation, voting prohibition, ARIS AI Companion use on OneDelaware process maps, and Benny chatbot launch are stated in the report. Companion governor press release same day confirms the report release and $117M estimated efficiencies framing without the AI legislative detail.
- Limitations and uncertainty
- $117M figure is estimated lifetime savings, not audited cash savings. AI legislative status is a Commission recommendation, not enacted law. No confirmation of document production to House Oversight. Benny and ARIS AI uses lack accuracy or time-saved metrics in the report.
Tamil Nadu signs MoU with IIT Madras Pravartak to stand up a PMU for an AI-ready state data centre
On 5 November 2026 the Tamil Nadu government, through ELCOT, signed a Memorandum of Understanding with the IITM Pravartak Technologies Foundation (IIT Madras) to create a Project Monitoring Unit that will oversee establishment of an Artificial Intelligence Data Centre for the state. The PMU's stated objectives are to evaluate the existing State Data Centre, prepare government applications to be "AI-ready," and plan migration to a new AI-ready facility. The signing at the Secretariat was attended by the Minister for Artificial Intelligence, Information Technology and Digital Services and the Additional Chief Secretary for that department. The article reports no budget figure, no construction timeline, no capacity metrics and no application migration results.
Why it matters, evidence and limitations
- Why it matters
- International state-level evidence that AI scale-up is being preceded by data-centre readiness and application migration planning. Transferable as a sequencing lesson for U.S. states modernizing shared hosting before enterprise GenAI; jurisdictional limits apply (Indian state procurement, ELCOT structure, IIT partnership model differ from U.S. state IT).
- Evidence and measured results
- Hindu Business Line article updated 5 November 2026 9:43 PM describes the Thursday MoU between ELCOT and IITM Pravartak, PMU objectives (evaluate existing SDC; make apps AI-ready; plan migration), and attending officials. No independent confirmation of facility specs in the opened page.
- Limitations and uncertainty
- MoU announcement only. No cost, schedule, capacity or success criteria published in the source. Indian state institutional arrangements are not 1:1 with U.S. states.
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.
- Government evaluation
- A public body’s measured evaluation or documented pilot.