Lighthouse AdvisorySLED AI Adoption Intelligence

Higher Education & ResearchIssue 174 records

College Athletics

Non-archived sources extend college-athletics AI coverage beyond the 2026-09-22 Zenodo/Seattle/Identity/NO FAKES set. Pitt CIC (2026-02-25) documents a production-path GenAI Survey Analysis Agent for Pitt Athletics fan feedback—RAG/semantic search on Amazon Bedrock with cited response IDs and open-source AWS stack—addressing slow dashboard cycles without published accuracy audits. KTRE (2026-08-10) reports SFA’s FanWord Assist adoption at $250–$350 per sport alongside student-journalist pushback on campus AI double standards, clarifying human-edit limits of box-score AI. FOIAball (2026-02-19) publishes Deloitte’s pre-settlement NILGo pitch language turning athlete inputs into Impact/Institution/Sport numerical scores, while the College Sports Commission’s June 23, 2026 memo confirms the live Deloitte-partnered Range of Compensation model, July prediction-interval refinement, and raised low-dollar RoC-review thresholds—together a control-layer pair on algorithmic NIL deal clearance.

Evidence records
4
Cross-source patterns
2
Evidence classes
3 independent reporting1 standards or public-body guidance
Outcomes
2 emerging1 mixed1 cautionary
Source freshness
3 older, newly relevant1 recent
Research completed
2026-09-24

Choose a role to see its takeaway beside every record in the ledger.

Patterns across the evidence

Synthesis · Lighthouse Advisory interpretation

2 patterns, each supported by at least two sources

  1. Athletic-department GenAI drafting tools expand coverage only when human publish/citation gates stay mandatory

    Pitt’s cited survey agent and SFA’s FanWord Assist both aim to speed athletics staff insight/content, yet each inspected source foregrounds human verification—response-ID citations at Pitt and explicit human editing plus live-play limits at SFA—while SFA also surfaces student-workforce tension. Together they show mid-major/DI ops AI value is contingent on governed draft lanes, not autonomous publishing.

    Operating questionWhere should athletic departments hard-require human sign-off before GenAI outputs affect public content or executive decisions?

    Supporting evidenceUniversity of Pittsburgh Cloud Innovation Center / Pitt AthleticsStephen F. Austin Athletics / FanWord (reported by KTRE)

  2. National NIL Go algorithmic RoC scoring forces local explainability and aggregate-threshold operations

    FOIAball’s Deloitte pitch-deck reporting shows Impact/Institution/Sport numerical scoring and AI pattern learning envisioned for NIL deal ranges, while the CSC’s June 23 memo confirms the live Deloitte-partnered RoC pillars, July prediction-interval refinement, and raised Associated-deal review thresholds. Schools cannot see weights but must still staff athlete communication, holistic-evidence packs, and ledger triggers.

    Operating questionHow should compliance teams operationalize opaque national RoC outcomes without building conflicting shadow FMV models?

    Supporting evidenceFOIAball / Deloitte (RFP proposals to NCAA)College Sports Commission

Evidence ledger

Full record · every source keeps its link and limitations

4 records
  1. Independent reportingEmergingNewly relevant · Feb 2026

    Pitt CIC builds GenAI Survey Analysis Agent so Athletics can query fan feedback in plain English

    Pitt CIC student developers, working with Pitt Athletics, built an open-source GenAI Survey Analysis Agent that lets staff ask natural-language questions over fan survey data. The system embeds survey text, applies similarity thresholds and keyword filters before LLM calls, uses RAG/semantic search via Amazon Bedrock, and returns summaries with citations tied to original response IDs. Associate AD Richard Turnquist describes replacing slow report/dashboard pull cycles. Supporting GitHub docs specify Titan embeddings in S3 Vectors, Claude Sonnet 4.5, and a serverless AWS stack. No independent accuracy or time-savings study is published in the success story.

    University of Pittsburgh Cloud Innovation Center / Pitt AthleticsUnited States (NCAA Division I; University of Pittsburgh)February 25, 2026

    Why it matters, evidence and limitations
    Why it matters
    Newly relevant non-archived DI athletic-department GenAI implementation focused on fan-ops qualitative data—complements archived Equipe chat and Morgan academic chatbot with an open AWS RAG pattern and explicit citation design.
    Evidence and measured results
    Pitt Digital article (2026-02-25 datetime): Turnquist quote on slow answers; Farrell similarity-threshold quote; Shelke citation/response-ID language; Bedrock integration; GitHub README stack and ~0.15 threshold testing.
    Limitations and uncertainty
    Success-story framing; no published hallucination audit; earlier 2026 date with why-now as first archive hit for this pattern; fan-ops scope not student-athlete performance AI.
    University success story — athletics GenAI survey analytics implementation Role takeaways & implementation
  2. Independent reportingMixedRecent

    SFA adopts FanWord Assist AI game recaps at $250–$350/sport amid student-journalist pushback

    KTRE reports Stephen F. Austin Athletics signed a multi-year FanWord Assist deal so a small communications staff can turn box scores, stats, and postgame interviews into editable recaps, bios, and award nominations. Assistant AD Jacob Pumphrey prices the tool at $250–$350 per sport—about one-eighth of a full-time staff cost—stresses human editing, and notes AI cannot observe live plays. A student journalist criticizes lost resume-building opportunities and a campus double standard discouraging AI in coursework while athletics deploys AI. SFA’s August 5 release confirms Assist features and a >215-school network claim without pricing or critique.

    Stephen F. Austin Athletics / FanWord (reported by KTRE)United States (NCAA Division I; Southland Conference; Nacogdoches, Texas)August 10, 2026 (SFA athletics release August 5, 2026)

    Why it matters, evidence and limitations
    Why it matters
    Newly relevant non-archived FanWord deployment with local scrutiny and unit-cost transparency—extends archived St. Thomas FanWord signal with mixed workforce evidence and explicit human-edit limits.
    Evidence and measured results
    KTRE (2026-08-10): pricing quote; human-element and live-play limitation quotes; Chandler critique; Southland soccer poll use. SFA release (2026-08-05): Assist capability list and network claim.
    Limitations and uncertainty
    Local reporting; pricing self-reported; one student voice; no content-quality audit; vendor network size unverified here.
    Independent local reporting — athletics AI content tool with workforce critique Role takeaways & implementation
  3. Independent reportingCautionaryNewly relevant · Feb 2026

    FOIAball publishes Deloitte NILGo pitch: Impact/Institution/Sport scores feed AI deal ranges

    FOIAball published Deloitte and PwC proposals from the NIL clearinghouse RFP that became NILGo. Deloitte’s deck describes evaluating third-party NIL deals by converting athlete and market inputs into numerical scores across Impact, Institution, and Sport categories into a Student Athlete Score used for fair-market ranges, including social followership/engagement, institutional reach, sport market size, and on-field performance percentiles, plus Payor alignment. The article highlights AI learning from retroactively flagged non-compliant deals. PwC’s losing bid (~$6M) was more explicit about automated approval bands around a specified FMV. Author cautions decks predate the House settlement and CSC standup, so production NIL Go may differ.

    FOIAball / Deloitte (RFP proposals to NCAA)United States (NCAA Division I / College Sports Commission NIL Go)February 19, 2026

    Why it matters, evidence and limitations
    Why it matters
    Newly relevant non-archived transparency on the algorithmic NIL clearinghouse every DI program touches—pairs with CSC’s live RoC controls and contrasts prior deepfake/NO FAKES coverage with deal-scoring AI.
    Evidence and measured results
    FOIAball (2026-02-19): Impact/Institution/Sport framing; numerical scores language; Student Athlete Score; social API crawl description; performance percentile examples; AI training-on-flagged-deals language; PwC $6M bid; pre-settlement caveat.
    Limitations and uncertainty
    Pre-settlement proposals; production model may have changed; investigative/advocacy tone; not a CSC-authored model card.
    Independent investigative reporting — NILGo algorithm pitch documents Role takeaways & implementation
  4. Standards or public-body guidanceEmergingNewly relevant · Jun 2026

    CSC memo: Deloitte RoC pillars, $50k Associated threshold, shift to prediction intervals

    The College Sports Commission’s June 23, 2026 memo states the Deloitte-partnered RoC model evaluates Associated NIL deals across institution market reach, social media reach, and student-athlete performance pillars; clearing within the RoC of at least one pillar can suffice, else holistic tests apply. Effective July 1, 2026, deals valued $600–$15,000 are not RoC-reviewed unless/until an athlete exceeds $50,000 in Associated deals in an academic year. CSC also approved moving the model from confidence intervals to prediction intervals in early July to better reflect compensation variability, with temporary forbearance on borderline over-bound deals during the update. The memo separately warns against institution–agent consulting payments that may circumvent rules.

    College Sports CommissionUnited States (NCAA Division I institutions and conferences)June 23, 2026

    Why it matters, evidence and limitations
    Why it matters
    Newly relevant official control-layer documentation of the live NIL Go scoring regime—anchors FOIAball’s pitch-deck scrutiny with current enforcement thresholds and methodological change, not previously archived.
    Evidence and measured results
    CSC PDF memo (2026-06-23): three pillars; one-pillar clearance rule; July 1 threshold change with examples; quarterly batching; prediction-interval language; agent consulting enforcement note.
    Limitations and uncertainty
    Does not publish weights or fairness audits; prediction-interval details limited; not an independent evaluation of model bias.
    Standards/guidance — official NIL Go RoC policy update memorandum Role takeaways & implementation

How to read this edition

Source findings, measured results and limitations come from the cited publications. Patterns, operating questions, role takeaways and implementation considerations are Lighthouse Advisory interpretation, stated as questions to validate locally rather than guaranteed outcomes. Vendor and operator claims are labeled as claims. Full research method.

Independent reporting
Independent reporting with attributable sources but without a formal evaluation design.
Standards or public-body guidance
Normative or advisory guidance from a standards body or public institution.