> **Status note added 5 September 2026.** This is the external audit of **v1.0**, kept as a record of what
> it found and when. Findings 1 through 6 were applied across v1.0.1 to v1.4: the inclusion-rule violations
> were moved to supplementary, the missing reports were added, the UNC date was corrected, the crawler
> metrics were split into requests attempted, pages read, AI candidates and other document links, the seed
> lists and raw crawl outputs were published, and the census was re-run twice. **Every row and census count
> below describes v1.0** (69 inventory rows, 187 census rows), not the current release, which has 105
> inventory rows and a 372-row census. In particular, the "roughly 55 to 60%" recall estimate on this page is
> the figure the project later **measured at 47%**; see the codebook.

# AI task force inventory: evidence audit

Audit date: September 5, 2026. Target: https://kylesaunders.com/ai-taskforce-inventory/

This audit identifies actionable errors and omissions in the live v1.0 inventory. It is a structural audit of all 69 inventory rows and 187 census rows, an HTTP retrieval attempt for every inventory URL, document metadata inspection, and a targeted external search. It is **not a completed full-text recoding of every recommendation or an exhaustive new search of all 187 institutions**. An unflagged row has not been certified correct. No changes were made to the live site.

## Findings that should be addressed first

1. Apply the published inclusion rule consistently. Dartmouth is a single-author special-adviser report; Ohio State's Fisher document is a slide deck. Both formats are expressly excluded by the Methods tab.
2. Add the missing UNLV and UNMC reports; add Michigan's missing 2023 and 2026 reports. These are different from the Michigan research report already present.
3. Fix the UNC–Chapel Hill document date: August 2, 2024, not September 2024.
4. Separate document existence, public availability, document genre, and institutional coverage. A public abstract is not a public deliberative report; a systemwide report is not exclusively a flagship-campus report.
5. Repair the crawl metrics before interpreting negative cases: the published crawler counts attempted requests, not successful pages inspected, and does not fetch the document links it counts.
6. Reconcile the codebook with the dataset and remove stale statements and placeholders.

## Existing cases: errors, conflicts, and adjudications

| Case | Finding and evidence | Recommended action | Confidence |
|---|---|---|---|
| Dartmouth College | The linked report's title page identifies James E. Dobson, Special Advisor to the Provost for AI, as author, June 30, 2025. The Methods tab expressly excludes single-author adviser memoranda. | Exclude under the current rule, or explicitly expand eligibility to commissioned adviser reports and apply that expansion to all institutions. Preserve the document in a supplementary category. | Confirmed rule conflict |
| Ohio State University, Fisher College of Business | The linked 53-page PDF is a slide presentation, including co-presenters, numbered slides and a closing Q&A. First page visually inspected. The Methods tab excludes slide decks. | Exclude from the report corpus under the present rule; keep as a presentation record. Do not infer that Ohio State has no other qualifying report. | Confirmed rule conflict |
| UNC–Chapel Hill | The linked PDF begins “Chancellor’s Generative AI Working Group / FINAL REPORT / August 2, 2024.” Current date_published is 2024-09. | Change document date to 2024-08, or 2024-08-02 if supporting daily precision. Preserve a separate posting date if September is independently supported. | Confirmed error |
| Indiana University | The March 2024 report describes the University Faculty Council and IU administration working across all IU campuses. Inventory assigns it exclusively to Bloomington UnitID 151351; Indianapolis is coded no public report surfaced. | Record systemwide scope and relate the report to covered campuses. Distinguish campus-authored from system-covered reports in the census. Do not duplicate the report as separate documents. | Confirmed scope mismatch; census treatment requires a rule |
| University of Chicago, both rows | Both linked PDFs are one-page announcements of a report's existence. Dataset correctly labels them abstract_only, but the institution is included in “Report found” and report KPIs. Neither page supplies findings or recommendations. | Retain as report-exists/public-abstract-only records; exclude from a strict public-deliberative-text numerator. Amend Methods, which mentions summary_only but not this exception. | Confirmed measurement inconsistency |
| Chicago, 2025 education report | Its key_recommendation claims a roadmap with baseline rules and departmental divergence. The linked 77-word abstract contains no such recommendation. | Remove that substantive claim or attach the additional source supporting it. An abstract can establish existence, not the report's unprovided recommendations. | Confirmed unsupported-by-linked-source claim |
| University of Colorado Denver / Anschutz, three rows | The teaching summary identifies CU Denver; the August 26, 2026 report describes CU Denver faculty from its seven schools/colleges and Auraria Library. The historical IPEDS entity label does not itself establish coverage of Anschutz. | Preserve UnitID for the join but add issuing_campus/coverage fields; identify the source institution as CU Denver unless Anschutz participation is evidenced. | Confirmed attribution ambiguity |
| CU Denver, August 2026 report | The report calls itself “Abridged report,” while doc_status is final. Other public summaries are coded summary_only. | Use separate fields for publication stage (final/interim) and completeness (full/abridged/abstract), or consistently classify abridged text. | Confirmed coding inconsistency |
| University of Rochester, education | The source is a guidelines page. It states original publication December 2, 2024, with revisions March 7 and May 16, 2025. Inventory records only 2024-12. | Add original and revision dates and a version identifier. Adjudicate whether committee-authored guidance qualifies under the exclusion of policy/guidance pages. | Confirmed version omission; eligibility requires adjudication |
| Rochester research; LSU; Kentucky | These are committee-associated guidance/recommendation documents. A blanket exclusion of guidance conflicts with inclusion based on committee provenance; LSU includes an executive summary and analytical context. | Define a substantive genre test: evidence/findings and recommendations from a named body, regardless of whether the title says guidelines. Avoid deciding by title alone. | Boundary cases, not automatic removals |
| Texas A&M | Direct automated fetch returned 404, but the same URL is linked from the current AI homepage and web retrieval exposed a 10-page strategy action plan. | Do not delete or declare broken on the basis of the 404 alone. Record access-method discrepancy and retain the official landing page. | Confirmed conflicting retrieval outcomes |
| UC systemwide | Cover letter is January 5, 2026, but the attached report itself is dated December 10, 2025 and was endorsed December 17. | Keep 2025-12 under the stated document-date rule. Do not change it simply because the transmittal date is 2026. | Existing date supported |
| Virginia Tech | Filename refers to September 2025, but the report title page says July 2025. Sponsors are the Provost and Chief Operating Officer. | Keep 2025-07. Consider multiple-sponsor support; current single provost code loses the COO's role. | Existing date supported |
| RIT | Included in inventory but absent from census, consistent with being outside the R1 frame. Its UnitID is blank. | Preserve as an explicitly supplementary non-R1 entry and populate its verified identifier separately. Clarify why non-R1 entries are included if the codebook describes an R1-only corpus. | Confirmed scope/metadata issue |

Source links for the case findings:

- [Dartmouth report](https://provost.dartmouth.edu/sites/provost/files/2025-12/Report%20on%20Artificial%20Intelligence%20at%20Dartmouth%20(Dobson,%202025.06.30).pdf)
- [Ohio State Fisher presentation](https://ai.osu.edu/sites/default/files/documents/Fisher%20COB%20Task%20Force%20on%20GenAI%205.6.2024.pdf)
- [UNC report](https://chancellor.unc.edu/wp-content/uploads/sites/1276/2024/09/UNC-Chapel-Hill-Generative-AI-Working-Group-Report.pdf)
- [Indiana University report](https://ufc.iu.edu/committees/IU-Gen-AI-Taskforce-Final-Report-3.27.24.pdf)
- [Chicago 2024 abstract](https://provost.uchicago.edu/sites/default/files/2024-10/AdvancingComputingandAIReport.pdf) and [2025 abstract](https://provost.uchicago.edu/sites/default/files/2025-08/AIEducationWGReport.pdf)
- [CU Denver teaching summary](https://www.ucdenver.edu/docs/aiatcudenverlibraries/default-document-library/t-l-exec-summary_ai-strategy.pdf?sfvrsn=b7dcfb4_1)
- [CU Denver August 2026 abridged report](https://www.ucdenver.edu/docs/aiatcudenverlibraries/administration-and-operations-working-group/82626-summer-2026-ai-faculty-liaison-cop-report---final.pdf?sfvrsn=9f1c9bb4_3)
- [Rochester education guidelines](https://www.rochester.edu/provost/gen-ai-education/)
- [Texas A&M official AI homepage](https://ai.tamu.edu/index.html)
- [UC systemwide report](https://senate.universityofcalifornia.edu/_files/reports/council-chair-to-senate-divisions-senate-ai-workgroup-report.pdf)
- [Virginia Tech report](https://ai.vt.edu/content/dam/ai_vt_edu/Responsible-and-Ethical-AI-Framework-for-Virginia-Tech-(v1.0-Sep-2025).pdf)
- [RIT classification explanation](https://www.rit.edu/news/new-carnegie-classifications-affirm-rits-reputation-science-and-technology)

## Missing reports within institutions already in the R1 census

| Institution | Missing document | Evidence / date | Census implication |
|---|---|---|---|
| University of Nebraska Medical Center | UNMC AI Task Force Report | [Institutional repository](https://digitalcommons.unmc.edu/unmc_reports/1/) identifies authors and **November 29, 2023** publication date, offers a full-text download, and describes five subcommittee investigations and recommendations. A [May 2024 university announcement](https://www.unmc.edu/newsroom/2024/05/28/task-force-report-outlines-future-for-ai-use-at-unmc/) independently confirms it. Full PDF retrieval was blocked in this session; existence, provenance, date, and substantive report character are documented by the repository. | Replace “Crawl too shallow to say” with report found, with repository verification noted. Do not use the 2024 news date as the report date. |
| University of Nevada, Las Vegas | Faculty Senate AI Task Force Final Report, May 2024 | [Official IT resource hub](https://www.it.unlv.edu/ai-technology/resource-hub) expressly links the final report and describes its completed charges and guidance. Full linked Drive document was not inspected in this audit. | Replace “Crawled; no public AI report surfaced” with report located, pending full-text extraction for substantive coding. Current report existence is well evidenced. |
| University of Michigan | Generative Artificial Intelligence Advisory Committee report, June 2023 | [Official committee reports page](https://genai.umich.edu/committee-reports) identifies and links the June 2023 report. It is distinct from the August 2024 research report already listed. | Add a report row; institution already positive. |
| University of Michigan | AI in Education Working Group report, April 30, 2026 document date | [Official reports page](https://genai.umich.edu/committee-reports); [67-page report](https://provost.umich.edu/wp-content/uploads/ai-in-education-working-group-report-k9X7mP2qL4vN8wR1zT5bY0cX3vM6nB9zQ2wE5rT8yU1iO4pA7sD0fG3hJ6.pdf). Cover date is April 30; retrieval title also references September 2, 2026, so preserve version/public-posting uncertainty. Co-sponsored by the Provost and VPIT-CIO. | Add a report row; institution already positive. Treat as a current addition, without asserting it was publicly accessible before the original harvest. |

The UNLV October 2024 “second report” link is labeled a presentation. Do not automatically add it as another qualifying report while slide decks remain excluded. Its February–March 2024 report is a further candidate, distinct from the May final report, requiring a version/duplication decision.

### Penn State: confirmed additional report listings; document inspection still required

The [Senate committee page](https://senate.psu.edu/senators/committees/other-committees/joint-standing-committee-on-effective-and-responsible-use-of-artificial-intelligence-in-higher-education-instruction/) lists material absent from the inventory:

- AI-Assisted Grading Report, Academic Year 2025–2026.
- AI-Enabled Assistive Technologies at Penn State: Accessibility, Integrity, and Compliance Report.
- 2025 Report from the Special Taskforce on AI-Enabled Assistive Technologies.

The [November 18, 2025 plenary agenda](https://senate.psu.edu/senators/senate-resources/senate-meeting-agendas-records/november-18-2025-plenary-agenda/) also links a Report from the Joint Standing Committee for AI in Instruction. A [faculty publication list](https://behrend.psu.edu/person/tiffany-petricini) corroborates the 2025 reports. These links currently point to SharePoint. I inspected the institutional listings, not the SharePoint documents. Determine whether the assistive-technology links are distinct reports or versions before assigning a count; confirm public access before calling them public full texts.

## Additional material for the supplementary inventory

These are not automatically errors in an R1-only census. The codebook says non-R1/system/international material was deliberately parked for later. Because the live inventory already includes UC systemwide and RIT, the scope should nevertheless be explicit and consistent.

| Institution / system | Document | Verification and recommended treatment |
|---|---|---|
| University of Nebraska system | AI Taskforce Report and Recommendations | [Full institutional HTML report](https://nebraska.edu/offices/provosts-office/academic-affairs/nu-ai-taskforce). Recommends a system AI institute and campus centers. Add as systemwide; do not attribute solely to Nebraska–Lincoln. Date needs verification. |
| Alfred University | AI Task Force Final Report, August 13, 2025 | [16-page PDF](https://my.alfred.edu/teaching-learning-center/_docs/ai-task-force-final-report-082025.pdf). Named task force, subcommittees, findings and recommendations verified. Add to non-R1 inventory. |
| Western Illinois University | An Initial Guide to Generative AI at WIU: The Fall 2023 AI Task Force’s Final Report to Faculty Senate | [26-page PDF](https://www.wiu.edu/faculty_senate/committees/reports/AI%20Committee%20Final%20Report%201-23-24.pdf). Cover date **December 29, 2023**, despite January 2024 filename. Add to non-R1 inventory. |
| Northern Illinois University | Task Force on AI in Academic Affairs: summary report of working-group findings, 2024–2025 | [Official page](https://www.niu.edu/artificial-intelligence/task-force/index.shtml) links [DOCX summary](https://www.niu.edu/provost/_files/niu-ai-task-force-summary-report-2024-2025.docx). Official description establishes deliberation and recommendations. Verify exact document date; add as a summary if eligible. |
| University of San Francisco, Gleeson Library | Gleeson Library AI Taskforce Final Report, 2025 | [Repository](https://repository.usfca.edu/gli/1/) identifies named authors, report type, deliberative process and recommendations. College/unit scope. Add to supplementary inventory after full-text check. |
| California State University system | CSU Generative AI Committee Report, June 2024 | [Official committee listing](https://genai.calstate.edu/about/systemwide-generative-ai-advisory-committee); [system explanation](https://www.calstate.edu/csu-system/faculty-staff/labor-and-employee-relations/Pages/budget-and-labor-agreement-faq.aspx) confirms its provenance and role in institutional decisions. Listing found in search; direct report not inspected. Add to acquisition queue. |
| University of Toronto | AI Task Force final report, June 2025 | [Official task force page](https://ai.utoronto.ca/u-of-t-ai-task-force/) confirms delivery and links report page. Add to international acquisition queue; outside U.S. census. |

## Unresolved cases that should not be called confirmed public-report additions

- **Colorado State:** The inventory codebook explicitly says a CSU report exists and that the initial verified-absence judgment was wrong, but supplies no report citation. The census still gives only an archive/no-AI note. The [CSU joint task force page](https://ai.colostate.edu/artificial-intelligence-joint-task-force/) verifies a system task force, named leadership, and a December 2024 recommendations milestone; it does not expose a report link. Reconcile the codebook claim by identifying the document and its access status. I have not independently located that report.
- **Northern Arizona:** Its [provost page](https://in.nau.edu/provost/academic-operations/resources-policies/artificial-intelligence/) explicitly refers to a June 2023 Generative AI in Education Task Force report but does not link the report. Code as report referenced, text not located; not as verified absence.
- **UConn CLAS:** The [December 2025 announcement listing](https://clas.uconn.edu/category/announcements/) confirms implementation of CLAS AI Task Force recommendations. Underlying report not obtained; news is not the report.
- **Rutgers–New Brunswick:** A [June 15, 2023 Cyberinfrastructure and Data Science task force report](https://newbrunswick.rutgers.edu/sites/default/files/2024-03/nb_cyberinfrastructure-and-data-science-task-force-report.pdf) recommends an institute for AI/data science/cyberinfrastructure. Potential addition if broad computing strategy qualifies, as it does for Chicago. Full eligibility not adjudicated here.
- **University of Idaho:** The [president's working-groups page](https://www.uidaho.edu/leadership/president/working-groups) confirms an AI working group established in 2024. No report recovered in this pass.

## Crawler and measurement audit

Source: [published crawl2.py](https://kylesaunders.com/ai-taskforce-inventory/crawl2.py).

### The displayed page count is not successful coverage

The crawler executes `n += len(batch)` before checking whether each fetch returned HTML. `get()` returns None for HTTP failures and non-HTML responses, and it suppresses all exceptions. Thus failed requests, missing seed paths, and non-HTML responses count toward `pages`. The output cannot by itself justify the UI label “governance pages crawled.” Homepages and non-governance routes may also be counted. A high value such as 152 is not evidence of 152 successfully inspected governance pages.

**Fix:** separate requests_attempted, html_pages_successfully_read, document_links_discovered, documents_successfully_read, HTTP status/error counts, and content-level classifications. Retain the URL-level log. Batch size eight can overshoot the 150 cap, explaining 152 without a data error.

### “Other task force documents reached” is overstated

Document URLs enter `hits` based on URL/anchor scoring. The crawler does not fetch those documents. The returned `other_taskforce_docs` count therefore means candidate document links, not inspected or necessarily valid task-force documents.

**Fix:** rename to non_ai_candidate_document_links, or retrieve and classify each document before reporting a verified-document count. Withdraw claims that large counts establish the absence of AI reports from an archive.

### Off-domain documents are discarded

Before recording a link, the crawler requires its hostname to belong to the institutional root. Officially linked Drive, Box, SharePoint and external CDN documents can therefore be dropped. UNLV demonstrates why this matters.

**Fix:** preserve outgoing document candidates linked from trusted institutional pages, with provenance. Offsite document collection does not require unrestricted offsite crawling.

### HTML reports are traversed but not emitted as report candidates

Only URLs matching PDF/DOC/PPT extensions are added to `hits`; potentially relevant HTML pages are queued but never emitted as candidate reports. This is mismatched with the inventory's own inclusion of HTML reports.

**Fix:** emit qualifying HTML report candidates as well as downloadable documents, then review genre and named-body provenance.

### Retrieval channels introduce differential missingness

Library/repository hosts are often excluded or not seeded; governance-looking subdomains are privileged; external document hosts are excluded. These decisions interact with campus web architecture. Aggregate known-positive recall of roughly 55–60% is not a subgroup-specific estimate and cannot correct private/public or established/new-R1 comparisons by itself.

**Fix:** describe the observed outcome as public reports found by a specified protocol. Report retrieval failures and source types by subgroup. Keep crawl status separate from report availability. Do not infer governance effort or institutional silence from non-discovery.

### Reproducibility is incomplete

The crawler requires r1_list.json with domain/name/unitid mappings. That input and the transformation producing census statuses and displayed statistics are not linked in Data & Downloads. The published crawler also does not retain the information needed to distinguish a WAF block from a 404 or other failure.

**Fix:** publish the actual seed file, raw crawl outputs, classification/adjudication log, and script generating the inventory/census/statistics. A census CSV alone is not a replacement for the crawler's input and transformation history.

## Website and codebook consistency

Sources: [inventory](https://kylesaunders.com/ai-taskforce-inventory/) and [codebook](https://kylesaunders.com/ai-taskforce-inventory/codebook.html).

- Live data contains **69 rows, 57 distinct institution labels, and 55 R1 census positives**. RIT and UC systemwide account for the two labels outside the census. Label the 57 figure so readers do not mistake it for 57 R1 campuses.
- Census status totals recomputed: report found 55; no report surfaced 108; blocked 11; too shallow 12; domain mapping failure 1. Total 187. Methods prose says seven institutions block automated access; census and codebook say eleven. If seven refers to a different pass, label that pass.
- The search no-results message says the census covers 146 institutions classified R1 in 2021. It actually contains 187 institutions under the 2025 frame.
- Find Your Institution searches only census rows, so RIT cannot be found there even though it has an inventory entry. Either explicitly label that search R1 Census or search the union with appropriate scope labels.
- The About tab retains “[N] rows at random myself.” Replace it with a documented count and sampling method, or remove the unperformed claim.
- The codebook says there will be no independent second coder but later promises a double-coded pass and an inter-rater statistic. Choose and state one actual validation design. A genuinely independent overlapping human/AI coding sample can support an agreement estimate if described appropriately; merely checking or editing the AI's answers is not independent coding.
- The codebook says IPEDS UnitID is not populated although almost all inventory rows and all census rows have IDs.
- The codebook's convening-authority counts are stale. Live values: provost 35; president 12; faculty_senate 7; joint_admin_senate 4; college_unit 3; research_office 3; cio_it 2; unknown 2; system_senate 1.
- The codebook publishes trial-code counts while the Methods tab says coding is withheld. Remove stale counts or clearly label them as abandoned, unvalidated trial outputs.
- The codebook's “Harvard has none” wording contradicts its own discoverability caveat. Use “no qualifying public report located under this protocol.”
- The codebook cites a parked OTHER-SEGMENTS.md file, but Data & Downloads does not link it. Link the supplementary inventory or remove the inaccessible reference.
- Separate stage from access/completeness: final/interim/recurring/implementation and full/summary/abstract/gated measure different properties and should not compete in one doc_status field.

## Practical update sequence

1. Preserve this v1.0 snapshot.
2. Correct UNC's date and unsupported Chicago recommendation; resolve Dartmouth/Ohio State and guidance genre rules.
3. Add UNMC/UNLV and the two Michigan reports; queue the Penn State documents for access and deduplication checks.
4. Add campus/system coverage relationships, maintaining one row per report and a separate many-to-many report–institution mapping.
5. Repair crawler metrics and candidate handling, then rerun negatives with URL-level evidence retained.
6. Recompute public-full-report, public-summary, public-abstract and institution totals separately.
7. Regenerate codebook counts and UI language from the same data; document a reproducible human spot-check.

A corrected grand total is intentionally not asserted here: exclusions, system coverage, abstract-only treatment, and Penn State version deduplication must first be resolved consistently.

## Appendix: retrieval record for all 69 inventory rows

These are initial direct-fetch outcomes from this audit, not certification of substantive correctness or persistent public availability. HTTP 200 can be a landing page or access interstitial; 403 can be an automated-access block. Texas A&M's initial 404 was contradicted by successful web retrieval of the same report.

| Row | Institution | Report | Initial HTTP | Audit retrieval detail |
|---|---|---|---|---|
| 1 | Boston College | [Student Education and Empowerment in the Age of AI](https://cdil.bc.edu/resources/emerging-technologies/engaging-with-ai/student-ai-advisors/ai-advisors-report-2025/) | 200 | HTTP check only; no PDF text extracted |
| 2 | Boston University | [BU AI Task Force Report on Generative AI in Education and Research](https://www.bu.edu/cds-faculty/files/2024/12/Boston-University-AI-Task-Force-Report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 3 | Brown University | [GAITL Committee Final Report and Recommendations](https://provost.brown.edu/sites/default/files/GAITL_Committee_Report_FNL.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 4 | Cornell University | [Generative Artificial Intelligence for Education and Pedagogy](https://teaching.cornell.edu/sites/default/files/2023-08/Cornell-GenerativeAIForEducation-Report_2.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 5 | Cornell University | [Generative AI in Academic Research: Perspectives and Cultural Norms](https://it.cornell.edu/sites/default/files/itc-drupal10-files/Generative%20AI%20in%20Research_%20Cornell%20Task%20Force%20Report-Dec2023.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 6 | Cornell University | [Cornell Generative AI in Administration Task Force Report](https://it.cornell.edu/sites/default/files/itc-drupal10-files/Cornell%20Generative%20AI%20in%20Administration%20Task%20Force%20Report-Jan2024.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 7 | Dartmouth College | [Artificial Intelligence at Dartmouth: Opportunities and Recommendations](https://provost.dartmouth.edu/sites/provost/files/2025-12/Report%20on%20Artificial%20Intelligence%20at%20Dartmouth%20(Dobson,%202025.06.30).pdf) | 200 | Single-author adviser report confirmed |
| 8 | Duke University | [Report to the Duke University Provost: AI at Duke Steering Committee](https://ai.duke.edu/wp-content/uploads/sites/101/2026/04/AI-at-Duke-Report-Public-Summary.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 9 | George Washington University | [AI@GW Mapping Report](https://strategicframework.gwu.edu/sites/g/files/zaxdzs7071/files/2026-07/PRES-2526-588_AI@GW_Mapping_Report-FINAL.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 10 | Indiana University Bloomington | [Indiana University Generative Artificial Intelligence (GenAI) Task Force Report](https://ufc.iu.edu/committees/IU-Gen-AI-Taskforce-Final-Report-3.27.24.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 11 | Kansas State University | [Building a Strategic Framework for Artificial Intelligence - AI Task Force Report](https://www.k-state.edu/provost/initiatives/artificial-intelligence/stewardship/docs/ai-task-force-report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 12 | Lehigh University | [Report and Recommendations from Lehigh's Generative AI Advisory Group](https://provost.lehigh.edu/sites/provost.lehigh.edu/files/Lehigh-Gen%20AI%20Advisory%20Group%20Report%205-9-2025.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 13 | Louisiana State University | [Faculty Senate Generative AI Committee Guidelines on the Use of Generative AI in Curriculum](https://www.lsu.edu/senate/standing-committees/gai_committee/2025_2026ay/gai_guidelines_fall2025.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 14 | Massachusetts Institute of Technology | [Report of MIT's Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training](https://aiandeducation.mit.edu/report/) | 200 | HTTP check only; no PDF text extracted |
| 15 | Michigan Technological University | [Final Report and Recommendations, MTU AI Working Group 2023-2024](https://www.mtu.edu/provost/ai/about/reports/final-report-and-recommendations-4-22-2024.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 16 | Mississippi State University | [Report on Generative Artificial Intelligence (GAI) and Instruction](https://www.provost.msstate.edu/sites/www.provost.msstate.edu/files/2023-11/AI%20Working%20Group%20Report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 17 | Northeastern University | [Final Report from the Information Technology Policy Committee, AY 2023-24](https://faculty.northeastern.edu/senate/wp-content/uploads/sites/2/2024/03/Final-Report-ITPC-.docx) | 200 | HTTP check only; no PDF text extracted |
| 18 | Ohio State University | [Generative AI in the Fisher College of Business](https://ai.osu.edu/sites/default/files/documents/Fisher%20COB%20Task%20Force%20on%20GenAI%205.6.2024.pdf) | 200 | Slide deck confirmed, including visual inspection |
| 19 | Pennsylvania State University | [2024 Report of the Joint Standing Committee on Effective and Responsible Use of Artificial Intelligence in Higher Education Instruction](https://ai.psu.edu/resources/2024-report-of-the-joint-standing-committee-on-effective-and-responsible-use-of-artificial-intelligence-in-higher-education-instruction) | 200 | HTTP check only; no PDF text extracted |
| 20 | Rochester Institute of Technology | [RIT AI Task Force Final Report](https://www.rit.edu/provost/sites/rit.edu.provost/files/docs/AI_Task_Force_Final_Report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 21 | Stanford University | [Report of the AI at Stanford Advisory Committee](https://provost.stanford.edu/sites/g/files/sbiybj29391/files/media/file/ai-at-stanford-report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 22 | Texas A&M University | [Advancing AI Leadership for Texas and the Nation (AI Strategy Action Plan)](https://ai.tamu.edu/_files/_documents/ai-report-final-2026.pdf) | 404 | Initial 404; same URL successfully retrieved by web tool as a 10-page PDF |
| 23 | Tufts University | [Shaping the Future of AI Education at Tufts](https://tufts.box.com/s/1mis67n0ro78pkv88de9bi6hlwvjsr05) | 200 | HTTP check only; no PDF text extracted |
| 24 | Tulane University | [AI in Pedagogical Practice: Executive Summary and Recommendations](https://ai.tulane.edu/sites/default/files/2024-07/AI%20in%20Classroom%20Committee%20Report%20Executive%20Summary.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 25 | University at Buffalo (SUNY) | [Report of the Task Force on Generative AI in Teaching and Learning](https://www.buffalo.edu/content/dam/www/academicaffairs-new/Report%20of%20theTask%20Force%20on%20Generative%20AI%20in%20Teaching%20and%20Learning_FINAL_ACCESSIBLE%20(8).pdf) | 200 | PDF text extracted; not a full substantive recode |
| 26 | University at Buffalo (SUNY) | [AI Task Force Implementation Report](https://www.buffalo.edu/catt/ai-resources/ai-task-force-implementation-report.html) | 200 | HTTP check only; no PDF text extracted |
| 27 | University of Alabama at Birmingham | [A Report on AI and the UAB Community](https://www.uab.edu/provost/images/images//initiatives/Report_-_AI_and_the_UAB_Community.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 28 | University of California (systemwide) | [Academic Senate AI Workgroup Report and Recommendations](https://senate.universityofcalifornia.edu/_files/reports/council-chair-to-senate-divisions-senate-ai-workgroup-report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 29 | University of California, Davis | [Advancing AI at UC Davis: Implementation Plan, Governance Framework, and Recommendations](https://leadership.ucdavis.edu/sites/g/files/dgvnsk1166/files/media/documents/AISteeringCommitteeReportFinal.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 30 | University of California, Santa Cruz | [Supplemental Report on the Use of AI for Research](https://senate.ucsc.edu/files/2026/06/COR_Supplemental-Report-on-AI-in-Research_20260615.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 31 | University of Chicago | [Report of the Advancing Computing and AI Committee](https://provost.uchicago.edu/sites/default/files/2024-10/AdvancingComputingandAIReport.pdf) | 200 | One-page public abstract, not full report |
| 32 | University of Chicago | [Report from the Artificial Intelligence and Education Working Group](https://provost.uchicago.edu/sites/default/files/2025-08/AIEducationWGReport.pdf) | 200 | One-page public abstract, not full report |
| 33 | University of Colorado Denver / Anschutz | [Executive Summary: AI Strategy Teaching and Learning Working Group](https://www.ucdenver.edu/docs/aiatcudenverlibraries/default-document-library/t-l-exec-summary_ai-strategy.pdf?sfvrsn=b7dcfb4_1) | 200 | PDF text extracted; not a full substantive recode |
| 34 | University of Colorado Denver / Anschutz | [AI Research Working Group: SWOT/SOAR Analyses, Use Cases, Policy Recommendations](https://www.ucdenver.edu/docs/aiatcudenverlibraries/research-working-group/research_final_deliverable.pdf?sfvrsn=fd76cfb4_1) | 200 | PDF text extracted; not a full substantive recode |
| 35 | University of Colorado Denver / Anschutz | [AI Faculty Liaison Community of Practice, Summer 2026: Abridged Report](https://www.ucdenver.edu/docs/aiatcudenverlibraries/administration-and-operations-working-group/82626-summer-2026-ai-faculty-liaison-cop-report---final.pdf?sfvrsn=9f1c9bb4_3) | 200 | PDF text extracted; not a full substantive recode |
| 36 | University of Florida | [UF AI BlueSky Taskforce Full Report](https://www.ai.ufl.edu/media/aiufledu/resources/UF-AI-BlueSky-Task-Force-Full-Report-April-2025.docx) | 200 | HTTP check only; no PDF text extracted |
| 37 | University of Kentucky | [UK ADVANCE Fall 2023 Guidelines and Recommendations for Generative AI in Instructional Contexts](https://ombud.uky.edu/sites/default/files/2023-08/GenerativeAI_GuidelinesAndRecommendations.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 38 | University of Maryland, College Park | [Report of the University of Maryland AI Commission](https://umd-pres.files.svdcdn.com/production/files/president/UMD_A.I.CommissionReport.pdf?dm=1755892170) | 200 | PDF text extracted; not a full substantive recode |
| 39 | University of Massachusetts Amherst | [Special Report of the Joint Task Force on Generative AI (Sen. Doc. No. 24-087)](https://www.umass.edu/senate/sites/default/files/2025-05/24-087%20JTFGAI%20-%20Final%20Report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 40 | University of Michigan | [Artificial Intelligence Research Committee Recommendations Report](https://research.umich.edu/wp-content/uploads/2024/11/AI-Report-2024.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 41 | University of Minnesota | [Presidential AI Task Force Report: Education Working Group](https://president.umn.edu/presidential-initiatives/completed-reports) | 200 | HTTP check only; no PDF text extracted |
| 42 | University of Minnesota | [Presidential AI Task Force Report: Research Working Group](https://conservancy.umn.edu/items/30db9ae4-d748-4cd5-86dd-073721836edb) | 403 | HTTP check only; no PDF text extracted |
| 43 | University of Minnesota | [Presidential AI Task Force Report: Administrative Operations Working Group Recommendations](https://conservancy.umn.edu/items/2ef7fb8c-930f-43d1-9c54-53f23654e772) | 403 | HTTP check only; no PDF text extracted |
| 44 | University of Missouri | [Report of the Task Force on Artificial Intelligence in the Learning Environment](https://provost.missouri.edu/wp-content/uploads/2024/08/TF-Report-AI-in-the-Learning-Environment.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 45 | University of New Hampshire | [Spring 2025 Report of the Special Committee for Artificial Intelligence](https://www.unh.edu/fac-senate/sites/default/files/media/2025-05/AI%20Committee%20Report%20-%20Spring%202025.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 46 | University of North Carolina at Chapel Hill | [UNC-Chapel Hill Generative AI Working Group Report](https://chancellor.unc.edu/wp-content/uploads/sites/1276/2024/09/UNC-Chapel-Hill-Generative-AI-Working-Group-Report.pdf) | 200 | Document date correction confirmed: 2024-08-02 |
| 47 | University of North Carolina at Charlotte | [AI Faculty Task Force Report to the Provost](https://aihub.charlotte.edu/wp-content/uploads/sites/44/2026/07/AI-Faculty-Task-Force-Report-to-the-Provost-August-2025-1.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 48 | University of Notre Dame | [Generative AI Task Force Final Report and Recommendations](https://ai.nd.edu/assets/566975/gait_report_and_recommendations.pdf) | 403 | HTTP check only; no PDF text extracted |
| 49 | University of Oklahoma | [AI Working Groups Final Report](https://www.ou.edu/content/dam/ai/documents/wgs-final-report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 50 | University of Pennsylvania | [High Standards for Critical Reading, Writing and Research in the Age of AI](https://pricelab.sas.upenn.edu/sites/default/files/CriticalApproachestoAIWhitePaper.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 51 | University of Pittsburgh | [Report of the Ad Hoc Committee on Generative AI in Research and Education](https://teaching.pitt.edu/wp-content/uploads/2024/04/Report-Ad-Hoc-Committee-on-GenAI-3-1-24.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 52 | University of Rhode Island | [Exploring the Potential of Generative AI at URI: Impact, Opportunities and Recommendations](https://drive.google.com/file/d/1nHBvYjnQwKRwPKf4oPqex3cbG7Ui5xKX/view) | 200 | HTTP check only; no PDF text extracted |
| 53 | University of Rhode Island | [University of Rhode Island Academic Affairs AI Task Force Report: 2026 Update](https://drive.google.com/file/d/1mR0FNQimpr1x6EGZpR9oXICf-2USKsI8/view) | 200 | HTTP check only; no PDF text extracted |
| 54 | University of Rochester | [Generative AI Use in Education](https://www.rochester.edu/provost/gen-ai-education/) | 200 | HTTP check only; no PDF text extracted |
| 55 | University of Rochester | [Responsible Use of Generative Artificial Intelligence in Research](https://www.rochester.edu/university-research/wp-content/uploads/2024/12/Research-and-GenAI-Guidlines_FINAL_12.12.24.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 56 | University of South Carolina | [Report of the 2024-2025 Provost's Task Force on the Use of Artificial Intelligence (AI) Tools in Teaching and Learning](https://sc.edu/about/offices_and_divisions/cte/teaching_resources/docs/ai_task_force_report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 57 | University of Southern California | [Committee on Information Services Final Report 2025-2026](https://academicsenate.usc.edu/wp-content/uploads/sites/6/2026/07/Committee-on-Information-Services-Final-Report2025-2026.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 58 | University of Texas at El Paso | [Final Report: Findings and Recommendations](https://www.utep.edu/ai/_files/docs/ai-task-force-report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 59 | University of Texas at San Antonio | [AI Task Force Report](https://provost.utsa.edu/initiatives/artificial-intelligence/docs/2026-ai-task-force-report.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 60 | University of Virginia | [Generative AI has Arrived. What Does this Mean for Teaching and Learning at UVA?](https://genai.provost.virginia.edu/task-force-report) | 200 | HTTP check only; no PDF text extracted |
| 61 | University of Washington | [Recommendations from the AI Task Force: A Vision for the Future of Education, Research, Work and Community](https://cdn.uw.edu/wp-content/uploads/sites/11/2024/10/02150921/AI-Task-Force-recommendations-1.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 62 | University of Washington | [AI Task Force Recommendations - Second Report](https://www.ai.uw.edu/wp-content/uploads/2025/07/FINAL-Second-AI-Task-Force-Report-2025-03-14-1.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 63 | University of Wisconsin-Milwaukee | [Artificial Intelligence Taskforce Final Report](https://uwm.edu/ai/taskforce/artificial-intelligence-taskforce-final-report/) | 200 | HTTP check only; no PDF text extracted |
| 64 | University of Wyoming | [Artificial Intelligence Chatbots Working Group: Recommendations to President Seidel](https://www.uwyo.edu/acadaffairs/_files/docs/final-report-aigw-february-2023.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 65 | Virginia Tech | [Responsible and Ethical Artificial Intelligence Framework for Virginia Tech](https://ai.vt.edu/content/dam/ai_vt_edu/Responsible-and-Ethical-AI-Framework-for-Virginia-Tech-(v1.0-Sep-2025).pdf) | 200 | PDF text extracted; not a full substantive recode |
| 66 | Washington State University | [Final Report from the WSU AI Strategy and Policy Task Force](https://provost.wsu.edu/documents/2023/09/summer-2023-ai-task-force-report.pdf) | 403 | HTTP check only; no PDF text extracted |
| 67 | Wayne State University | [Ad hoc Subcommittee on AI Running Report - Winter 2024](https://academicsenate.wayne.edu/informational_reports/ad-hoc-subcommittee-on-ai-running-report-final-v2.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 68 | Wayne State University | [Ad hoc Subcommittee on AI Running Report - Fall 2025 & Winter 2026](https://academicsenate.wayne.edu/ai/ad-hoc-subcommittee-on-ai-running-report-25-26-final.pdf) | 200 | PDF text extracted; not a full substantive recode |
| 69 | Yale University | [Report of the Yale Task Force on Artificial Intelligence: Reflections and Recommendations](https://provost.yale.edu/sites/default/files/files/Yale%20Task%20Force%20on%20AI%20Report%20FINAL.pdf) | 200 | PDF text extracted; not a full substantive recode |