AI is reshaping higher education. What should the future of sports education look like?
Norway’s AI commission proposes changes to learning, skills and assessment. NST examines implications for sports science, coach education and active health — and why adapting today’s model is not the only question.
<p><strong>Norway’s Malthe-Sørenssen Commission wants to change what university students learn, how they learn and how their competence is assessed. Its report, NOU 2026:14, presented on 6 November, raises a question for sport: how should we educate the next generation of coaches, sports scientists and physical activity and health professionals?</strong></p> <p>This goes beyond allowing or banning a chatbot in an examination. When AI can contribute to analysis, training plans and technical explanations, education must develop both independent understanding and the ability to use technology critically.</p> <h2>The commission: New skills, more reliable assessment</h2> <p>In <a href="https://www.regjeringen.no/no/dokumenter/nou-2026-14/id3191187/">“Artificial intelligence in higher education: In the service of knowledge”</a>, the commission recommends regular, AI-informed reviews of curriculum content, learning activities and assessment. Students need general AI literacy alongside discipline-specific skills. Educators need the competence to support responsible use that promotes learning.</p> <p>It recommends greater use of supervised examinations and controlled components alongside submitted work. Short take-home examinations should be phased out. Substantial assignments should retain their educational role, but cannot alone establish what students know. Combined assessments should provide an overall judgement of learning outcomes, rather than merely adjust an assignment grade slightly.</p> <p>The report also identifies opportunities for AI-supported feedback and simulation, while warning that outsourcing thinking too early or uncritically can undermine learning. National computing and data-storage infrastructure, together with shared governance principles, should reduce vulnerabilities and supplier dependence.</p> <p>These are recommendations, not adopted policy. According to the <a href="https://www.regjeringen.no/no/aktuelt/utvalg-kunstig-intelligens-krever-nye-grep-i-hoyere-utdanning/id3191851/">Ministry of Education and Research’s announcement</a>, the report will go to consultation before the government considers further action.</p> <h2>NST’s submission: What if we started again?</h2> <p>On 2 September, Norwegian Sport Tech submitted <a href="https://nettsteder.regjeringen.no/wp-content/uploads/sites/589/ninja-forms/2/Norwegian-Sport-Tech_innspill_Malthe_Sorenssen_utvalget.pdf">“From adapting to AI to the future of higher education”</a>, published in the <a href="https://malthesorenssenutvalget.no/2026/09/02/norwegian-sport-tech/">commission’s public submissions archive</a>. NST is therefore both the publisher of this analysis and a participant in the consultation process.</p> <p>The submission made six proposals: explore education designed with AI as a starting point; give human–AI collaboration greater prominence; demonstrate competence both independently and with AI; trial new models across complete modules and programmes; respond more closely to changing workplace needs; and measure understanding and mastery, rather than simply the quality of submitted outputs.</p> <p>There is thematic overlap with the report, but no basis for claiming NST’s submission caused its recommendations. The commission explicitly addresses hybrid intelligence, including in section 7.8: both the outcome of student–AI collaboration and how the technology was used are relevant to assessment.</p> <p>NST’s further argument is therefore not that the commission overlooked this interaction. It is that education should also examine more fundamental organisational changes, alongside necessary adjustments to existing models.</p> <h2>From movement data to a coach’s judgement</h2> <p>For the Norwegian School of Sport Sciences, NIH, and other institutions teaching sport and health disciplines, this raises practical educational questions. The examples below represent NST’s analysis, not those institutions’ adopted plans.</p> <p>Teaching in biomechanics, exercise physiology, performance analysis and sports technology could connect physiological measurements and movement data with critical evaluation of AI-based analysis. Students need to understand measurement error, sampling and model limitations, rather than simply produce a convincing chart or explanation.</p> <p>In coach education, the distinction between data, analysis and responsibility is essential. Decision-support systems may suggest patterns or workload adjustments. Coaches must still evaluate the evidence, the athlete’s circumstances and professional appropriateness. Trust, communication and human relationships cannot be reduced to a model’s recommendation.</p> <h2>More practice — and an assessment that withstands questions</h2> <p>AI-supported simulation and adaptive exercises could potentially provide more practice scenarios: an athlete with conflicting workload signals, a recreational participant struggling with motivation or a rehabilitation process departing from its plan. Individual feedback might support progression, but a plausible simulation does not itself demonstrate improved learning.</p> <p>One possible assessment would ask students to develop a training programme using physiological data and AI assistance, document their use of the tools and verify the suggestions. They would then defend their decisions without AI: why this workload? What supports the assumptions? What would make you change the plan?</p> <p>This is a proposed experiment, not a requirement imposed by the report. It could make both subject understanding and competent technology use visible. Practical observation and professional supervision would still have a place.</p> <h2>Active health needs competence, not simply more tools</h2> <p>In prevention, rehabilitation and inclusion, courses could teach graduates to interpret activity data critically and adapt interventions to different circumstances. Potential AI services must be evaluated for privacy, bias and the risk of excluding people who do not fit the model. Generative AI is not a substitute for professional health judgement.</p> <p>The report provides an important foundation for change. NST also wants to bring three possibilities into discussion: pilot programmes in sport, movement and active health; assessment of both independent competence and human–AI collaboration; and partnerships between education, sport, research and technology companies to develop and evaluate learning models. These are not approved or funded projects.</p> <p>The objective should be to establish what students genuinely understand, retain and can apply over time. Should we simply teach future sport professionals to use AI, or also investigate how it could help them learn more, develop better judgement and prepare more effectively for the responsibilities ahead?</p> <p><small>Sources: NOU 2026:14, the ministry’s announcement of 6 November and NST’s submission of 2 September 2026, linked above. Checked on 7 November 2026. Sporting examples and possible next steps represent NST’s analysis. The hero is AI-generated and does not depict an actual teaching environment at NIH.</small></p>