The Complete Guide to Using AI in the Government Industry in Lawrence in 2025

Lawrence's government cannot treat AI as a distant trend - Kansas already requires statewide guardrails and practical controls, so local leaders must pair innovation with concrete data hygiene and workforce planning.

The Kansas Office of Information Technology Services generative AI policy requires agencies to prohibit State Restricted Use Information in GenAI queries and to disclose contractor AI use, while national guidance from Government Technology Insider and NASCIO highlights workforce upskilling, ethics reviews, and six adoption best practices to reduce risk and unlock value.

Lawrence's recent Innovations in GIS award shows local capacity to modernize data and workflows - so the immediate task is inventorying record series, hardening privacy controls, and training staff to use AI tools safely for faster, more equitable services.

“It is essential that we be proactive in finding the best way to use any technology that can pose risks to Kansans' data and privacy. With the adoption of this policy, Kansas serves as a model for what an enterprising, effective government can do to stay at the forefront of technological advancements.”

Generative AI (GAI) is a class of machine‑learning systems - large language models and multimodal engines like ChatGPT, Microsoft Copilot, and Google's Gemini - that produce text, images, or code from a user prompt by predicting word sequences from huge datasets; for practical government use this means tools can automate OCR, summarize long records, draft agenda notes, and suggest outreach language, accelerating routine workflow while leaving judgment to staff.

But these systems also hallucinate, invent fake sources or people, and reflect biases in their training data, so human review, source verification, and equity checks are essential before any AI output becomes a public record or policy brief; KU's teaching resources offer concrete classroom‑style approaches to probe outputs and adapt assignments to AI, while their bias guidance lays out common categories of dataset and cultural skew to watch for.

For a grounded start, review foundational definitions and technical capabilities, pilot narrow tasks like document extraction and customer‑service summaries, require verification steps, and train teams to critique model outputs rather than accept them at face value - doing so turns GAI from a risky black box into a dependable productivity partner for Lawrence's government.

Learn more in KU Center for Teaching Excellence generative AI and bias primer, and Google AI technical overview: what AI does.

“It makes mistakes. It makes things up.” - KU Center for Teaching Excellence

The 2025 industry outlook shows AI moving from experiment to core infrastructure, a shift Lawrence must treat as an operational priority: the U.S. AI market is estimated at USD 173.56 billion in 2025 with a long‑term CAGR near 19% (U.S. AI market forecast), private AI investment hit roughly $109.1 billion in 2024 and generative AI continues to attract major capital and performance gains, while inference costs have fallen dramatically - making advanced models far more affordable for local governments.

That combination means tangible upside for Lawrence: targeted pilots can yield the 20–30% productivity gains PwC highlights if paired with governance and workforce reskilling, but capture of those gains depends on disciplined data hygiene, procurement controls, and phased value-driven projects.

Use the Stanford HAI 2025 AI Index to benchmark risks and sector trends and lean on the PwC playbook to prioritize high-impact pilots that pair municipal data with clear oversight and measurable ROI (Stanford HAI 2025 AI Index report, PwC 2025 AI business predictions).

“Top performing companies will move from chasing AI use cases to using AI to fulfill business strategy.” - Dan Priest, PwC US Chief AI Officer

Federal AI policy in 2025 remains fragmented - built on the National Artificial Intelligence Initiative Act, executive orders, and agency guidance - while states are where concrete obligations land, so Lawrence must treat state rules as operational reality rather than distant policy.

All 50 states introduced AI bills in 2025 and, according to the National Conference of State Legislatures, 38 states adopted or enacted roughly 100 measures this year, with actions ranging from disclosure and automated‑decision inventories to ownership and deepfake limits (NCSL 2025 state AI legislation summary and tracker).

U.S. regulators and legal commentators note there is no single federal AI Act yet, just a mix of executive direction and sectoral rules, and recent executive orders (including January 23, 2025 actions) shift priorities at the national level - so local governments should expect changing federal guidance alongside binding state laws (U.S. AI regulatory tracker and analysis, Overview of 2025 U.S. AI legislation and state actions).

Practical takeaway for Lawrence: prioritize an automated‑decision inventory, algorithmic impact assessments, and vendor contract clauses now - these steps map directly to the kinds of state requirements already being enacted and reduce the risk of retroactive compliance costs.

Kansas has moved from cautious study to practical guardrails: the Office of Information and Technology Services' statewide generative AI policy requires human review of AI outputs, bars State Restricted Use Information from GenAI prompts, treats inputs into tools like ChatGPT as part of the public record, and frames contracts to prevent vendors from mining citizen data (Kansas statewide generative AI policy and OITS guidance (StateScoop)); at the same time the legislature enacted HB 2313, which bans use of certain “platforms of concern” (including DeepSeek and models owned or controlled by specified foreign actors) on state‑issued devices and forces agencies to deactivate and delete prohibited accounts (Kansas HB 2313 ban on platforms of concern (Inside Government Contracts coverage)).

Local legal and professional practice adds more detail: Shawnee County's District Court Rule 3.125 requires lawyers to verify AI‑drafted pleadings and disclose AI use at filing, a model Lawrence should adopt for municipal filings and RFP reviews (Guidance on generative AI ethics and Shawnee County Rule 3.125 (Baker Sterchi)).

Practical takeaway for Lawrence: inventory restricted data, forbid GenAI inputs that include PII or restricted records, add vendor clauses that prohibit secondary use of municipal data, and treat AI outputs strictly as draft material requiring named human signoff before becoming an official record.

“We need positive control of the data to ensure we have privacy.” - Jeff Maxon, Kansas interim CIO/CISO

Federal agencies are not only experimenting with AI - they are integrating it into operations at scale, a reality with direct lessons for Lawrence: the Government Accountability Office found agency AI use cases nearly doubled (571 → 1,110) and generative AI submissions jumped roughly ninefold (32 → 282) from 2023 to 2024, illustrating both rapid adoption and attendant management gaps (GAO report on generative AI use and management); at the same time the federal AI Action Plan signals new funding and incentives that favor states adopting permissive, investment‑friendly rules, which means Kansas's regulatory posture will shape Lawrence's eligibility for federal infrastructure and workforce grants (Summary of America's AI Action Plan and federal incentives).

Practical takeaways: treat the GAO's inventory and risk‑management expectations as required checkboxes, map municipal use cases to OMB procurement and governance guidance, and prepare grant‑ready proposals that pair clear data governance with workforce training so Lawrence can capture federal investments rather than being penalized for unclear controls (Analysis of OMB memos and federal AI procurement guidance).

“reassert American leadership in artificial intelligence,”

Data privacy and records compliance are the gating items for any AI rollout in Lawrence: Kansas does not yet have a comprehensive state privacy law, so municipalities must rely on specific state IT policies, court rules, and federal sector rules while filling gaps locally; see a concise summary of Kansas's current statutory landscape (Kansas data protection guide - state privacy overview).

Practically, Kansas Rule 24 places the redaction burden squarely on filers - attorneys or unrepresented parties must remove personally identifiable information and a court clerk has no duty to review filings - so automated drafting or model-assisted pleadings should include human review and PII‑detection checkpoints before any submission (Kansas Rule 24: Protection of Personally Identifiable Information - court filing redaction requirements).

At the agency level, the ITEC-8010-P Kansas Data Review Board policy requires formal data governance roles, an annual roster to the state CITA, role‑based training within 30 days, and a maintained data inventory with classification, retention, and recovery objectives - controls that must be mirrored in any AI procurement and vendor contract language to prevent unauthorized secondary use (ITEC-8010-P: Kansas Data Review Board policy - data governance and inventory requirements).

So what: because clerks won't catch unredacted PII and state policy demands named data owners and 30‑day training, Lawrence must prioritize an intake checklist, automated PII scans, and contract clauses that ban reuse of municipal data before expanding AI pilots.

Successful AI adoption in Lawrence depends less on exotic tools than on people: build pathways that combine classroom leadership with on‑the‑job apprenticeship, expand civilianized roles that free sworn staff for higher‑value work, and create named data owners with fast, role‑based training so AI outputs are reviewed and auditable.

Start by aligning city hiring and reskilling with statewide momentum - Kansas business leaders explicitly back new Youth Apprenticeship and Pre‑Apprenticeship pilot programs (Lawrence Chamber 2025 Legislative Priorities) - and use local training hubs like the Lawrence Workforce Center WIOA and training resources for WIOA funding, mock interviews, and computer lab access to upskill hourly and IT staff.

Pair that pipeline with leadership cohorts - KU's KU Leading EDGE high-performance program in Lawrence runs intensive, practical modules (next cohort July 18–25, 2025; fee $3,750) that help managers reconfigure structure, strategy, and incentives for distributed decision‑making.

One concrete metric to track: convert at least 20% of routine positions to trained civilian or tech‑assistant roles within 18 months to capture measurable productivity gains while protecting records and privacy.

“Our civilian staff are indispensable.” - Chief Rich Lockhart

Translate strategy into six concrete steps for Lawrence: 1) pick value‑first, narrow pilots tied to clear outcomes (start with document extraction, customer service summaries, or finance value‑chain work) as Government Technology Insider recommends; 2) inventory data and classify record series, ban State Restricted Use Information from model prompts, and run automated PII scans before any pilot; 3) harden security and governance - adopt Zero Trust principles and named data owners to meet NASCIO/Cloudflare 2025 priorities; 4) bake procurement and vendor clauses into every RFP to forbid secondary use of municipal data and require algorithmic impact assessments, using NASPO/NASCIO procurement guidance to shorten timelines and manage vendor risk; 5) pair pilots with role‑based training and a 90‑day upskill plan so staff can verify AI outputs (a concrete target: preserve auditability and human signoff on every AI‑drafted public record); and 6) measure and scale from pilots to operations using NASCIO's “AI Blueprint” checklist of governance considerations - this roadmap turns abstract policy into manageable tasks and can materially reduce approval cycles (one state CIO reported compressing procurement timelines from 18 months to six months when governance and procurement were aligned).

Use the linked NASCIO blueprint, procurement playbook, and state/local best practices to make each step auditable and grant‑ready for federal funding.

“AI has the power to transform state operations but requires coordinated efforts between CPOs and CIOs.”

Move from planning to short‑cycle execution: within 90 days Lawrence must complete a data inventory and automated PII scan, lock down State Restricted Use Information from any GenAI prompts, and add vendor clauses that forbid secondary use of municipal data so pilots stay grant‑ready as federal procurement priorities shift; pair that with a concrete workforce target - convert 20% of routine roles to trained civilian or tech‑assistant posts within 18 months - and enroll managers and staff in ethics and practical AI training (for example, KU's free CEU session on “Ethics, Technology, and AI in Social Work” and the statewide Kansas Digital Government Summit on Sept.

30, 2025) to build verification habits and procurement literacy. For hands‑on skills, channel staff into applied coursework like Nucamp's AI Essentials for Work to teach prompt design, verification workflows, and vendor oversight that make pilots auditable and scalable.

“Our digital world requires social workers who can navigate the technology landscape with integrity, ensuring that they uphold the dignity and values of the profession in both physical and virtual spaces.”