From the College Athletics edition of September 26, 2026
Frontiers experiment: AI labels raise credibility/diagnosticity; humans still win perceived athlete uniqueness
Inha University (Frontiers in Psychology) · Academic sport psychology / performance evaluation research · International (UK football-fan sample; Korean authorship; pro football stimulus)
- Publisher
- Frontiers in Psychology
- Original publication
- 10 September 2026
- Source retrieved
- 2026-09-27
What happened
Uhm et al. (published 2026-09-10) randomly assigned 220 UK football fans to identical Kevin De Bruyne performance reports labeled as AI-based or professional human analysts. AI attribution significantly increased message credibility and perceived diagnosticity, while human attribution increased perceived consideration of athlete uniqueness (partial eta-squared about 0.05-0.10). Authors stress perception is not accuracy and caution against generalizing to professional decision makers. Why now: same-week FanWord/performance-AI labeling wave makes source-disclosure design newly operational for college athletics.
Why it matters
Transferable caution for collegiate performance and communications teams presenting AI-assisted athlete evaluations—trust can inflate on systematic dimensions while athletes still perceive uniqueness neglect.
Evidence and measured results
Between-subjects experiment; MANCOVA/ANCOVA with familiarity covariate; adjusted means reported; sensitivity analysis including failed manipulation-check participants preserved direction.
Limitations and uncertainty
Pro football stimulus and fan sample—not NCAA coaches/athletes; content held constant so no objective accuracy comparison; effect sizes moderate/dimension-specific.
Put this evidence to work
Lighthouse Advisory interpretation, grounded in this source. Enriched 2026-09-27; this does not change the original publication date. Each role follows the same checklist so takeaways compare across sources.
Sales
Role takeaway
- Customer problem
- Departments adopting AI performance reports and athlete content risk miscalibrated trust when the AI label alone changes perceived credibility.
- Stakeholders
- Sports performance / analytics staff
- Head coaches
- Student-athletes
- Athletics communications
- Discovery questions
- How are AI-assisted evaluations currently labeled to coaches and athletes?
- Is human review disclosed, and who is accountable when a report is wrong?
- Which decisions (playing time, roster, rehab) may never rely on AI-only attribution?
- Value hypothesis
- Clear hybrid disclosure and uniqueness-context fields can reduce overtrust on systematic dimensions while addressing uniqueness-neglect concerns—without claiming better accuracy.
- Potential engagement
- Bounded label/design pilot with coaches on identical reports under AI, human, and hybrid attribution; measure reliance intentions and complaint themes.
- Unsupported claims
- AI reports are objectively more accurate than human analysts
- Findings generalize to NCAA coaches or scouts (fan sample)
- AI systems inherently consider athlete uniqueness
Pre-sales engineering
Role takeaway
- Fit
- Source-attribution and explanation UX for athlete performance reports rather than a predictive model purchase.
- Architecture and integration
- Report template fields for source label, data inputs, uncertainty, and human contextualization; audit log of who approved release.
- Prerequisites
- Defined report schema with provenance fields
- Human reviewer role in workflow
- Agreement on which audiences see which labels
- Constraints
- Study held content constant; effects are perceptual
- UK football-fan sample, professional football stimulus
- No objective accuracy benchmark against match truth
- Security
- Avoid overstating model capability in UI copy; keep sensitive athlete notes out of fan-facing labeled reports.
- Proof of value
- Coach reliance calibration tasks under alternate labels
- Athlete perception of uniqueness consideration
- Rate of inappropriate high-confidence decisions after AI-labeled reports
Delivery
Role takeaway
- Work
- Publish a source-disclosure standard for AI-assisted performance and communications outputs; train staff; run a small coach pilot.
- Dependencies
- Sports performance leadership
- Communications brand standards
- Existing report tooling
- Ownership
- Performance analytics owns report content; athletics leadership owns disclosure policy.
- Skills and adoption
- Teach staff that AI labels can raise credibility/diagnosticity perceptions without improving truth.
- Governance checkpoints
- Disclosure policy sign-off
- Pilot ethics/IRB-equivalent review if athletes are research subjects
- Quarterly review of mis-reliance incidents
- Acceptance criteria
- All AI-assisted athlete evaluations carry source and human-review fields
- Documented ban on presenting fan-study perceptions as accuracy proof
- Coach pilot completed with written trust-calibration findings
- Risks
- Overtrust leading to roster or rehab decisions on thin evidence
- Athlete pushback on uniqueness neglect
- Transferring pro-football fan findings uncritically into NCAA ops
Implementation considerations
Lighthouse Advisory interpretation across the operating dimensions a public-sector buyer must settle before this evidence becomes a design. Each note answers the question under its heading for this specific source.
Architecture and integration
What must connect, and where does the AI sit in the workflow?
Invest in report UX: source labels, input provenance, uncertainty, and human contextualization fields—not only model scores.
Governance
Who approves, reviews and stays accountable for outcomes?
Disclosure policy and human-review statements are governance controls against overtrust; do not cite fan credibility lifts as proof of analytical quality.
Security and privacy
What data, permissions and controls need testing?
Limited direct privacy findings; practical risk is oversharing sensitive athlete evaluations under an AI-credibility halo.
Accessibility and workforce
Who is affected, and what skills or accommodations follow?
Staff must be trained that AI labels can raise perceived diagnosticity without improving truth—affecting how analysts present findings to coaches.
Procurement
What should contracts, pricing and exit terms secure?
When buying performance-report or storytelling AI, require disclosure/explainability features and hybrid-review workflows in acceptance tests.
Operating model
Which teams own the service once it runs?
Supports hybrid operating models where AI drafts systematic summaries and humans explicitly own individualized athlete context.
Publication history
- 2026-09-26College Athletics · Issue 205 resources
Stable resource ID: frontiers-uhm-ai-human-source-attribution-2026-09-10