{"resourceId":"pa-oor-2026-aoro-survey-ai-requests-redaction-20261019","versions":[{"version":"external-2fd3d1aaec7e4f32578706b3461f7781f8d7eca6031a0aeb234cecc55329b0a1","resource":{"id":"pa-oor-2026-aoro-survey-ai-requests-redaction-20261019","title":"Pennsylvania Office of Open Records survey: 36% of agencies saw AI in at least some requests, 39% could not tell, and only 31% use redaction software","organization":"Pennsylvania Office of Open Records (OOR)","sector":"State and local public records (Right-to-Know Law)","geography":"Pennsylvania, United States","publishedAt":"2026-10-19 (PDF creation date; no date printed in the report)","publicationDate":"2026-10-19","eventDate":null,"sourceName":"Pennsylvania Office of Open Records","sourceLabel":"OOR: 2026 Agency Open Records Officer Survey Report (PDF)","sourceUrl":"https://www.openrecords.pa.gov/Documents/2026_AORO_Survey_Report-Final.pdf","evidenceClass":"government-evaluation","outcomeClass":"cautionary","topics":["operating-model","governance-procurement","data-security"],"finding":"The OOR surveyed 4,066 registered agency open records officers and received 1,265 responses (about 31%); respondents reported 89,559 requests in 2025, which the OOR extrapolates to roughly 285,000 to 288,000 statewide. 36% said AI was used in at least some requests, 39% could not determine whether AI was used, and 25% observed none. Nearly 90% received at least some commercial requests and 55% said all or most requests were commercial. Only 31% use electronic redaction software. The OOR now sees over 500 appeals a month, up from a 2024 average of 269, and passed its 2025 total of 3,970 appeals by the end of September.","sledRelevance":"A statewide, quantified look at the public records workload that AI tools are targeting from both sides: requesters using AI and agencies still redacting by hand. Most respondents are townships, boroughs and school districts, so it describes small local agencies well. Other states' laws and fee rules differ.","evidence":"Self-reported survey with a 31% response rate; AI categories are respondent perceptions, and the report notes the high-volume agencies that skew the average while some large agencies did not respond. Appeal counts come from OOR's own docket.","architectureImplications":"With two-thirds of agencies apparently redacting manually, basic electronic redaction and request tracking come before any AI review layer.","governanceImplications":"The OOR suggests policies on fees for complex datasets, standards for automated filings and processing rules, within current law; it flags commercial requests for possible legislative scrutiny.","securityPrivacyImplications":"Manual redaction at scale raises the risk of releasing personal information; redaction tooling and review checks matter.","caveats":"Perception-based AI measures; partial response; one state; no date printed in the PDF (date taken from file metadata).","streamIds":["citizen-engagement"],"roles":{"sales":"The selling point here is basic capacity, not advanced AI. A state survey shows records officers seeing AI in requests while about 69% still lack electronic redaction software. For Pennsylvania locals, discovery should cover request volume, redaction method and appeal exposure. A bounded engagement could be a records workflow assessment or a shared-service model for small agencies. The survey does not measure any tool's benefit.","engineering":"Start from the baseline the survey describes: mostly small agencies, many fulfilling requests on paper, most without electronic redaction. Engineering fit is request tracking, electronic redaction with review, and secure delivery, then optional AI-assisted search for responsive records. Integration with email, document stores and paper scanning matters. Test redaction accuracy on real document samples before any claim of time savings.","delivery":"Delivery should plan staged improvements: logging every request, standard electronic redaction with second-person review, written policies on automated or commercial filings, and appeal tracking. Training is central because the OOR itself flags redaction tools for upcoming training. Acceptance targets should be set against a measured baseline of hours per request and appeal rates. These are proposals, not results observed in the survey."},"roleDetails":{"sales":{"customerProblem":"Records officers face rising AI-assisted and commercial request volume with mostly manual redaction.","stakeholders":["Agency open records officer","Solicitor or legal counsel","Township or borough manager","School district administrator"],"discoveryQuestions":["How many requests and appeals did you handle in 2025, and how many involved commercial requesters?","Do you redact electronically, and who checks redactions?","Do you have a written policy on automated or bulk filings?"],"valueHypothesis":"Electronic redaction and tracking could reduce manual effort and appeal risk; size of benefit is unmeasured here.","potentialEngagement":"A records workflow assessment for a group of small agencies, scoped to tracking, redaction and policy.","unsupportedClaims":["That AI requests caused the appeal spike on their own","Any time savings from redaction software","Results outside Pennsylvania"]},"engineering":{"fit":"Fits request tracking, electronic redaction and document search for small local agencies.","architectureIntegration":"Connect request intake, document repositories and redaction with a review step and an audit log.","prerequisites":["Scanned or digital records","Request log","Redaction review procedure"],"constraints":["Right-to-Know Law timelines and exemptions","Many agencies still fulfill on paper","Small budgets"],"security":"Redaction failures expose personal data; keep audit logs of who redacted and reviewed each release.","proofOfValue":["Measure redaction accuracy on a sample of past releases","Compare hours per request before and after on a pilot group"]},"delivery":{"work":"Introduce request tracking, electronic redaction with review, and automated-filing policy for small agencies.","dependencies":["Legal counsel","Records software vendor","County or regional shared-service sponsor"],"ownership":"Agency open records officer, with solicitor oversight.","skillsAdoption":"Staff need training on redaction tools and on handling commercial and automated requests.","governanceCheckpoints":["Policy review by counsel","Redaction quality check before go-live","Annual comparison with the next OOR survey"],"acceptanceCriteria":["Proposed: 100% of releases pass second-person redaction review","Proposed: hours per request tracked monthly against baseline"],"risks":["Redaction errors","Staff not adopting tools","Fee policies challenged on appeal"]}},"retrievedAt":"2026-10-31T03:03:15Z","enrichedAt":"2026-10-31T03:08:32Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Respondents describe requests taking more time and legal review, stretching limited staff; the OOR plans training on redaction tools.","procurementImplications":"Small agencies may need shared or state-provided redaction and tracking tools rather than individual purchases.","operatingModelImplications":"Records work is shifting toward handling commercial and automated volume; intake rules and fee policy are as important as tools.","sourceVerification":{"openedUrl":"https://www.openrecords.pa.gov/Documents/2026_AORO_Survey_Report-Final.pdf","referenceExcerpt":"Only 31% of responding agencies indicated that they currently use electronic redaction software.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}