AI for FOIA and Public Records Requests

Use AI agents to log, search, de-duplicate and organize records, while records officers make every exemption and release decision

A government course from ibl.ai (self-paced, 8 to 10 hours, free). A hands-on, self-paced course for records officers, FOIA staff, clerks and agency IT: apply AI agents to intake, search, de-duplication and review preparation for FOIA and state public records requests, while a records officer makes every withholding and release decision. Taught from 5 U.S.C. § 552 and DOJ, NARA and NIST guidance.

Last updated:

Course description

AI for FOIA and Public Records Requests teaches records officers, FOIA and public records staff, clerks, agency counsel and agency IT to use AI agents on the slow parts of a records request without moving any legal decision out of human hands. You measure where the working days go against the federal FOIA clock in 5 U.S.C. § 552 and your own state statute, design intake and routing an agent can run, document a search across record systems the agency can defend (including the prompts an agent used, as NARA's August 2026 AI guidance contemplates), de-duplicate a responsive set without deleting anything, configure exemption flags that leave foreseeable harm and every withholding to a records officer, measure flag precision and recall, keep a withholding log that supports the release letter, communicate with requesters in plain language and govern the pilot with the NIST AI Risk Management Framework. Eight labs on closed requests or synthetic data build the artefacts for a capstone workflow and pilot plan. Last verified 18 November 2026.

About this course

Records requests keep rising while staff and statutory deadlines do not move: the Department of Justice reports more than 1.5 million federal FOIA requests in fiscal year 2024, an increase of more than 110 percent since fiscal year 2015. This course applies AI agents to the slow, mechanical parts of a request (intake, search, de-duplication and review preparation) and keeps every legal decision with a records officer: what an agent may flag, what it must never decide, and how to keep a defensible record of review. It teaches the federal Freedom of Information Act from its text at 5 U.S.C. § 552, with the DOJ Guide to the FOIA, OIP and OGIS guidance, NARA records guidance and the NIST AI Risk Management Framework, and asks state and local learners to apply their own state statute throughout.

What you will be able to do

  1. Map your office's request lifecycle against the federal FOIA clock in 5 U.S.C. § 552(a)(6) and your own state's public records statute, and measure where the working days go using closed requests.
  2. Design request intake, triage and routing an agent can run, with every track, validity, fee and expedited processing decision left to a person.
  3. Document a search across record systems that the agency can defend, including the prompts an agent used and the systems it could not reach.
  4. De-duplicate and organize a responsive set without deleting anything, and record the reduction a reviewer can reproduce.
  5. Configure exemption flagging that leaves every withholding decision, including foreseeable harm, to a records officer, and measure flag precision and recall against real decisions.
  6. Keep a withholding log and record of review that supports the release letter and any appeal, communicate with requesters in plain language, and write a pilot plan governed with the NIST AI RMF.

Who this course is for

Roles: records officers, FOIA officers and public records coordinators, clerks, agency counsel who review withholdings, records managers, and agency IT staff who support search and redaction tools.

Agencies: city, county, state and federal agencies. Federal agencies answer to the FOIA; state and local agencies answer to their state public records law, which differs in deadlines, exemptions and fees.

Before you start: working knowledge of your agency's request process, your state statute if you are not federal, and access to a sample of closed requests. Python is optional: every lab can be done in a spreadsheet.

Format

Time: 8 to 10 hours, self-paced. Level: intermediate.

Hands-on: eight labs on closed requests or synthetic data, each producing an artefact you keep for the capstone.

Structure: a Start here section, four content sections of readings, worked examples, templates and labs, each ending in an AI assessment, and a capstone in which you design a records request workflow, process one closed request through it and write a pilot plan, followed by a reading list.

Grading: the four section AI assessments count for 60 percent of the grade and the capstone assessment for 40 percent. The pass mark is 70 percent.

Currency

Last verified 18 November 2026 against 5 U.S.C. § 552, the DOJ Guide to the FOIA (Procedural Requirements updated 15 August 2025), OIP guidance, NARA's August 2026 guidance on AI materials and General Records Schedule 4.2, the OGIS report on NARA's 2025 Records Management Self-Assessment, and NIST AI 100-1.

The rules are still changing. NIST states that the AI RMF 1.0 is being revised, NARA says it may issue further AI guidance, and courts and OIP revise FOIA guidance. State laws differ from the federal FOIA. Check the current sources before relying on any template. A records officer, with agency counsel, makes every withholding and release decision. This course is training, not legal advice.

Outline

  1. Section 0 · Start here

    • 0.1 Welcome and orientation
  2. Section 1 · Measure the request lifecycle and route requests an agent can run

    • 1.1 Where the time goes in a records request
    • 1.2 Intake, triage and routing
    • 1.3 Section 1 assessment
  3. Section 2 · Search the record systems and organize the responsive set

    • 2.1 A search the agency can defend
    • 2.2 De-duplicate and organize the responsive set
    • 2.3 Section 2 assessment
  4. Section 3 · Flag exemptions, decide, redact and keep a record of review

    • 3.1 An agent flags, a person decides
    • 3.2 Redact and keep a defensible record of review
    • 3.3 Section 3 assessment
  5. Section 4 · Communicate with requesters and run the end-to-end workflow

    • 4.1 Plain language, status and proactive disclosure
    • 4.2 The request-processing workflow
    • 4.3 Section 4 assessment
  6. Section 5 · Capstone and wrap-up

    • 5.1 Capstone: Records request workflow and pilot plan
    • 5.2 Capstone assessment
    • 5.3 Reading list and next steps