Higher Education & ResearchIssue 134 records
College Athletics
Newly archived academic and reporting sources deepen college-athletics AI coverage without repeating yesterday’s Sportico/SBJ/Fisher Phillips set. Claros et al. (Sports Medicine) report a single-institution NCAA post-concussion musculoskeletal injury ML model with AUC 0.82 and high precision at a low false-positive operating point—measured methods, not a deployed product study. Brewer et al. (Frontiers) document how DI coaching staffs actually use GPS, force plates, and related tools, including privacy pushback on continuous monitoring and capacity strain. Front Office Sports shows the adversarial side of recruiting AI: doctored highlight reels that staff can often spot but that still tax mid-major diligence. Temple’s NCAA Innovations–funded JournAI/Sam mentor extends athlete-support chatbots beyond Morgan’s academic-support grant into wellness/career referral—with only a small pilot behind claimed behavioral change.
- Evidence records
- 4
- Cross-source patterns
- 2
- Evidence classes
- 2 academic research2 independent reporting
- Outcomes
- 2 mixed1 cautionary1 emerging
- Source freshness
- 4 older, newly relevant
- Research completed
- 2026-09-20
Choose a role to see its takeaway beside every record in the ledger.
Patterns across the evidence
Synthesis · Lighthouse Advisory interpretation
Measured injury-risk ML outpaces coach capacity and continuous-monitoring consent
Claros shows a carefully specified post-concussion MSK risk model can reach strong held-out discrimination on standard concussion measures, while Brewer shows the same DI environments already struggle with tech workload and raise privacy concerns about 24-hour monitoring. Publishing a good AUC is not the same as operationalizing risk scores without staffing, athlete buy-in, and purpose limits.
Operating questionBefore deploying an injury-risk score to coaches, what staffing, consent, and decision-rights controls are required so the model does not become always-on surveillance?
Supporting evidenceUniversity of Delaware / Sports Medicine (Springer)University of Florida / Frontiers in Sports and Active Living
Recruiting and athlete-support AI still need human integrity rails
FOS shows AI can be used by recruits to game film review, forcing staff diligence and reputational sanctions, while Temple’s JournAI explicitly positions AI chat as a bridge to campus humans rather than a replacement. Across both lanes, AI changes throughput—but accountability stays with coaches, compliance, and student-affairs professionals.
Operating questionWhere must humans retain final say on recruiting authenticity judgments and athlete-support escalation when AI accelerates intake?
Supporting evidenceFront Office SportsBilly Penn at WHYY / Temple University
Evidence ledger
Full record · every source keeps its link and limitations
Claros et al.: Collegiate post-concussion MSK injury ML model reports AUC 0.82
Using 194 NCAA student-athletes (155/39 train/test) and 135 candidate variables from demographics, concussion characteristics, and standard clinical assessments, Claros et al. report a composite ML risk score with AUC 0.82; at a 6.67% false-positive rate, sensitivity was 79% and precision 95%. Top predictors included cognitive, balance, and reaction measures. Results support identifying elevated post-concussion MSK risk for targeted prevention—not a multi-site product deployment.
Why it matters, evidence and limitations
- Why it matters
- Direct collegiate sports-medicine evidence for athletic trainers and performance staff considering AI risk stratification after concussion, with explicit metrics and CARE/SCAT-aligned inputs transferable to other NCAA programs that already collect these assessments.
- Evidence and measured results
- PMC full text of Sports Medicine article (doi:10.1007/s40279-025-02196-4): methods (WoE transform, L1-penalized logistic selection, AIC), AUC/sensitivity/precision figures, Key Points, and single-institution CARE Consortium enrollment context.
- Limitations and uncertainty
- Single institution; test set n=39; no published multi-site external validation; not an evaluation of a commercial product; secondary press may overstate '95% accuracy.'
Brewer et al.: DI coaching staffs use sports tech amid workload and 24-h monitoring privacy concerns
Five focus groups with 17 DI coaching-staff members show GPS/IMU, force plates, and motion tools are embedded in practice when aligned with coaching goals, supporting decisions, individualized athlete management, and communication. Staff also face growing technology demands; privacy concerns about continuous or 24-hour monitoring appeared in three groups. Study is coach-side qualitative evidence, not an AI accuracy benchmark.
Why it matters, evidence and limitations
- Why it matters
- Explains the operating reality athletic departments face when layering AI/analytics on wearables and force-plate stacks—capacity, relational trust, and monitoring-scope limits—complementing archived athlete-interview CHI work with coach perspectives.
- Evidence and measured results
- PMC/Frontiers full text: methods (five focus groups; 17 staff; team sports listed), technology inventory categories, findings on decision support and communication, privacy concern count (n=3 groups), and limitations (single well-resourced DI site; no athlete voices).
- Limitations and uncertainty
- Single DI university; team sports only; coach perspectives only; qualitative themes are not measured injury or performance outcomes.
Front Office Sports: AI-doctored recruiting reels usually fail coach scrutiny but still burn staff time
Reporting describes high-school athletes using AI to assemble—and sometimes exaggerate—highlight reels. Washington State's Brad Larrondo cites isolated glitchy/robotic tells; full-game film review still catches fakes. FAU's Dan Cornely warns exposure can blackball a prospect across the coaching fraternity. Mid- and low-major staff face disproportionate diligence costs. AI film fraud is framed as integrity and workload risk, not undetectable deepfake success.
Why it matters, evidence and limitations
- Why it matters
- Complements archived Athletic coverage of staff-side AI film triage by documenting the adversarial misuse vector and why human full-game verification and reputation sanctions remain central to collegiate recruiting integrity.
- Evidence and measured results
- Opened FOS article (Sam Becker, Mar 18, 2026) with on-record quotes from Larrondo (Washington State), Cornely (FAU), and Starks (player development) describing detection tells, blackball consequences, and mid-major burden.
- Limitations and uncertainty
- Anecdotal prevalence; no measured detector accuracy; football-heavy; basketball impact speculative in-source.
Temple NCAA grant funds JournAI/Sam student-athlete wellness mentor chatbot
Temple researchers received roughly one-third of a $100,000 2025 NCAA Innovations in Research and Practice Grant to build JournAI/'Sam', an AI chat mentor for career planning, mental-health support, and referral to campus resources. A spring pilot with about 30 DI/DII/DIII athletes reportedly showed significant pre/post behavioral-change survey differences; the PI stresses the sample is small and more data are needed. Product framing explicitly supplements—not replaces—human support and is not for writing academic work.
Why it matters, evidence and limitations
- Why it matters
- Extends collegiate athlete-support AI beyond archived Morgan academic-success chatbot funding into psychosocial wellness/career referral with NCAA Innovations backing—useful for departments designing human-escalation rails.
- Evidence and measured results
- Opened Billy Penn report (Kariuki, Oct 9/21 2025) quoting Elizabeth Taylor on grant share, pilot size/diversity aims, claimed survey behavioral change, confidentiality aspirations, and supplement-not-replace positioning.
- Limitations and uncertainty
- Reporting on early pilot; effect sizes and safety evaluation not published here; self-reported survey change; production privacy unproven.
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.
- Academic research
- Research produced through an academic institution or peer-reviewed venue.
- Independent reporting
- Independent reporting with attributable sources but without a formal evaluation design.