Wirewalk Instituteof AI
pub-110 Basic AvailableFlagship

AI for Public Servants

For federal, state and local staff who use an approved AI tool on real public-sector tasks, from answering a constituent letter to working a FOIA release, and must get the records, privacy and fairness duties right.

Level
Basic
Study time
8 hours
Lessons
20 in 6 modules
Exam
20 questions

About this course

This is not a general tour of AI in government. Every module follows a named public servant through a task they actually do: a correspondence specialist drafting a reply to a resident about a utility bill, a federal regulatory analyst summarizing thousands of comments on a proposed rule, a FOIA officer scoring AI-suggested PII redactions, a policy analyst checking every citation in a researched briefing memo, a state eligibility worker auditing OCR-extracted pay stubs, a 311 representative handling escalations from a public-facing virtual agent, a contract specialist drafting market research and a performance work statement, an HR specialist writing a job announcement and structured interview questions, a clerk turning a council recording into minutes, a permit technician re-checking an AI plan pre-check, a grants specialist reviewing progress reports, and a child-welfare caseworker preparing for a home visit. For each one you see the task today, where AI helps, where it must not decide, the data and systems involved, the tool categories and named products in use reviewed in November 2026; current local configuration still requires verification (described from public sources, never endorsed), the rule that bites, and how to measure whether it worked.

The rules are taught where they apply, not in a separate legal module: OMB M-25-21 (workforce AI literacy, agency generative-AI policies, and the presumed high-impact categories for benefits, employment and translation), NARA's August 2026 guidance on AI prompts and outputs as federal records alongside state public-records acts, the Privacy Act (systems of records, SORNs, routine uses and PIAs), CUI handling, FOIA exemptions, the Plain Writing Act, Section 508 and the ADA Title II web rule, the FNS Advanced Automation memo for SNAP, CAPTA confidentiality, the Procurement Integrity Act, NIH's ban on generative AI in peer review, and state laws such as Texas's TRAIGA disclosure duty and HB 3512 training requirement. The course is built for the AI literacy that OMB M-25-21 expects of federal staff. Completing the required lessons, Decision Labs and automated exam earns a Wirewalk Institute completion-and-assessment certificate. It does not by itself satisfy any federal, state or agency training mandate (including Texas HB 3512, which requires a DIR-certified program), and it is not an accreditation or a demonstration of workplace competence; an automated multiple-choice exam measures knowledge of the lessons, not performance on the job. The course is educational, not legal advice. All exercises use synthetic data labeled SYNTHETIC — TRAINING ONLY. The core course ends with a synthetic cumulative Decision Lab and a one-page pilot-evidence note; workplace measurement is optional, which leads into pub-210.

What you will be able to do

  • Draft a constituent reply with an approved AI tool that removes unneeded PII, keeps Privacy Act and CUI data out of tools not covered for it, verifies every factual claim against the policy source, and identifies which prompts and outputs must be captured as records under NARA AC 11.2026 or a state public-records act
  • Rewrite a public notice into plain language and translate it so that every legal condition and critical field (dates, amounts, deadlines) survives, verified by a condition checklist and back-translation
  • Produce a public-comment summary, FOIA redaction pass or briefing memo in which every quote and citation traces to its source, seeded substantive comments or PII items are recalled, and the AI search prompts are documented for the case file
  • Audit AI-extracted fields and virtual-agent transcripts in a benefits or contact-center queue, route errors and urgent contacts to a human, and write the appeal-trace note the FNS automation memo expects
  • Decide, for a given task, whether AI output would be a principal basis for a benefits, employment or translation decision presumed high-impact under OMB M-25-21 section 6, and name the human review and appeal step that must remain
  • Prepare procurement, hiring, minutes or permit documents with AI while keeping proposal data, applicant ranking, unapproved minutes and permit decisions out of the tool, and correct planted errors in an AI draft
  • Design a measured pilot for one task in your own unit (baseline, accuracy or recall check, error and appeal rates) and write a one-page use-case note for your Chief AI Officer or IT office

Syllabus

6 modules · 20 lessons · about 8 hours

Module 1: The Correspondence Specialist: Answering a Constituent Letter With an Approved Tool

One reply to a resident about a utility-bill increase, used to teach the approved-tool rule, data minimization, claim checking and records capture that every later module relies on.
3 lessons
  1. Which Tool You May Use: Your Agency's Generative AI Policy, OMB M-25-21 and State Training Laws Such as Texas HB 3512
  2. Drafting the Utility-Bill Reply: Stripping Unneeded PII, Checking Every Claim Against the Policy Sheet and Routing for Clearance
  3. Your Chat Log Is a Record: NARA AC 11.2026, State Public-Records Acts and the Washington City ChatGPT-Log Requests
  • Decision Lab · 10 minThe Reply Due Before CouncilA synthetic Harlow Springs decision sequence with competing service strategies, a new evidence reveal and a scoped handoff.

Module 2: The Communications Specialist and Front-Line Staff: Plain-Language Notices, Alt Text and Translation

Turning a dense benefits notice into a plain-language, accessible page and a translated hearing notice without changing what the law says.
3 lessons
  1. Rewriting a Water-Assistance Notice in Plain Language Without Dropping a Legal Condition
  2. Drafting and Correcting Alt Text for Two Charts: Section 508, the ADA Title II Rule and the Manual Tests AI Cannot Replace
  3. Translating a Hearing Notice and a Resident's Message: Back-Translation, Critical Fields and M-25-21 Section 6(o)
  • Decision Lab · 10 minOne Notice, Three Ways to Miss a HearingA synthetic Harlow Springs decision sequence with competing service strategies, a new evidence reveal and a scoped handoff.

Module 3: The Regulatory Analyst, the FOIA Officer, the Grants Specialist and the Policy Analyst: Reading and Researching at Volume

Clustering public comments, scoring AI-suggested FOIA redactions, extracting grantee metrics and researching a briefing memo, with every output and citation traced back to its source.
4 lessons
  1. Summarizing Public Comments on a Proposed Rule: Form-Letter Dedup, Issue Clusters and Tracing Every Quote to a Comment ID
  2. Scoring AI-Suggested PII Redactions in a FOIA Email Release: Exemptions (b)(5), (b)(6) and (b)(7)(C) and the Search-Prompt Note
  3. Reviewing Grantee Progress Reports Against Award Targets, and Why NIH Bars Generative AI From Peer Review
  4. Drafting a Briefing Memo With a Web-Grounded Tool and Checking Every Citation Against the Primary Source
  • Decision Lab · 10 minThe Missing Paper-Form CommentA synthetic Harlow Springs decision sequence with competing service strategies, a new evidence reveal and a scoped handoff.

Module 4: The Eligibility Worker, the Contact-Center Agent and the Caseworker: Serving People Without Letting AI Decide

Benefits verification, virtual-agent escalations and child-welfare case preparation, taught from the side of the worker who keeps the decision.
4 lessons
  1. Working a SNAP Recertification Queue: Auditing OCR-Extracted Pay Stubs Under the FNS Advanced Automation Memo
  2. Handling Escalations From a Public-Facing Virtual Agent: AI Disclosure, Knowledge-Base Accuracy and Language Access
  3. Preparing for a Child-Welfare Home Visit From a Long Case File Inside the Approved System: CAPTA Confidentiality and the Missing Court Order
  4. The High-Impact Line From the Front Desk: Principal Basis, Human Review and Appeal Under M-25-21 Sections 4 and 6
  • Decision Lab · 10 minReceived Is Not ExtractedA synthetic Harlow Springs decision sequence with competing service strategies, a new evidence reveal and a scoped handoff.

Module 5: The Contract Specialist, the HR Specialist, the Clerk and the Permit Technician: Documents That Carry Legal Weight

Market research and a PWS, a job announcement and interview guide, council minutes and a zoning pre-check, each with the line AI must not cross.
4 lessons
  1. Market Research and a Draft Performance Work Statement Without Exposing Proposal Data: The Procurement Integrity Act and Your Agency's FAR Deviation
  2. Writing a Job Announcement and Structured Interview Questions That Match the Job Analysis and Leave Applicant Screening to People
  3. Council Minutes From a Recording: Checking Motions, Seconds, Vote Tallies and Public Commenters Before the Body Approves Them
  4. Re-Checking an AI Zoning Pre-Check at the Permit Counter and Writing the Correction Notice
  • Decision Lab · 10 minBefore the Council Packet Is Called FinalA synthetic Harlow Springs decision sequence with competing service strategies, a new evidence reveal and a scoped handoff.

Module 6: Capstone: Measuring One AI-Assisted Task in Your Own Unit

Picking one task from the course, measuring it honestly, and writing it up so your Chief AI Officer or IT office can act on it.
2 lessons
  1. Baselines and Measures That Matter: Turnaround, Clearance Edits, Recall on Seeded Items and Appeal Rates
  2. Writing a One-Page Use-Case Note for Your Chief AI Officer or IT Office
  • Decision Lab · 10 minThe Fast Pilot and the Two CorrectionsA synthetic Harlow Springs decision sequence with competing service strategies, a new evidence reveal and a scoped handoff.
  • Capstone simulation · 35 minThe Six-Week Briefing Under a Changed PermissionAssemble the module artifacts across a source revision, a records request and a changed data authorization before choosing a bounded next phase.

Exam and certificate

The exam unlocks once you have completed every lesson and played each Decision Lab to an ending. It draws 20 questions from a larger bank, spread across the modules, with the answer options shuffled for each attempt. You have 40 minutes and need 80% to pass. After a failed attempt you can try again after 24 hours, up to 3 attempts in total. Feedback is given per module, never as the answer key.

Passing earns a certificate of completion and assessment from the Wirewalk Institute of AI with a unique ID that anyone can check at institute.wirewalk.com/verify. It is not an accredited qualification or a vendor certification. How certification works.

The Wirewalk Institute of AI is independent. This course teaches publicly documented products and is not affiliated with or endorsed by GSA USAi, OpenAI, Anthropic, Google, ICF, RTI International, OPEXUS, Veritone, Granicus, Salesforce, CivicPlus, HeyGov, PublicInput, Archistar, Northwoods, or Microsoft. Product and company names are trademarks of their respective owners.

More in AI in Government, Defense and Public Safety

All 12 courses in this track
pub-001 Free intro Available

AI in Government, Defense and Public Safety: The One-Hour Briefing

A free one-hour briefing built on twelve real public-sector jobs, not an overview: summarizing public comments in an approved assistant, extracting SNAP verification documents, flagging PII for FOIA, reviewing an AI-drafted police report, escalating from a 911 virtual agent, handling a facial-recognition lead, checking an AI bench memo, drafting on GenAI.mil, acting on a predictive-maintenance alert, and drafting a grant. Each lesson shows where AI helps, where it must not decide, and the rule that governs that exact task.

1 h · 9 lessons Free intro
pub-210 Intermediate Waitlist

Building and Deploying AI in Government

How public bodies adopt AI as of October 2026: OMB M-25-21 and M-25-22 in practice, use case inventories, buying through GSA schedules, OneGov and FedRAMP 20x, service delivery use cases, due process in automated eligibility decisions, records and security obligations, and lessons from international public-sector frameworks.

12 h · 18 lessons Join the waitlist
pub-510 Executive Waitlist

AI Strategy, Risk and Regulation for Government Leaders

AI Strategy, Risk and Regulation for Government Leaders: an executive course in the AI in Government, Defense and Public Safety track. Syllabus in preparation.

h · 0 lessons Join the waitlist