Special Session 2 - Symposium on Ethical, Legal and Social Aspects of Artificial Intelligence

Sound governance is crucial to the pro-innovative, effective and ethical adoption of artificial intelligence (AI) technologies in healthcare. In Hong Kong, the Digital Policy Office and the Privacy Commissioner for Personal Data have issued relevant guidance on the use of AI in general, but there is as yet no explicit governance framework tailored to the medical context. 


This paper aims to provides an overview on AI-related regulatory regimes in other jurisdictions (e.g., Mainland China, the European Union, the United Kingdom, the United States, and Canada) with a view to illustrate the general trajectory of development and how existing regulatory characteristics may differ and evolve. It will explore whether some of the approaches, if any, may be applicable to Hong Kong, or whether a new regulatory mechanism needs to be established. It is the author's personal view that a dialogue on these matters is timely and necessary if Hong Kong is to assume a leading role in propelling the advancement of next-generation healthcare services.   

The application of AI in sports and exercise medicine is transforming the field by enhancing injury prevention, diagnosis, rehabilitation, and performance optimization. For injury Prevention & risk assessment, AI algorithms analyze motion data from wearables, cameras, or sensors to assess biomechanics, gait, or technique, identifying patterns linked to injury risks (e.g., ACL tears, stress fractures), as well can do predictive analytics with machine learning models process historical and realtime data (training load, sleep, biomarkers), to flag athletes at risk of overuse injuries. AI can also contribute in diagnosis & imaging, enhances interpretation of MRI, X-ray, or ultrasound scans, detecting subtle fractures, ligament tears, or muscle injuries with speed and accuracy, or even symptom analysis to triage symptoms and recommend specialist referrals. AI have also been used to design adaptive recovery protocols based on injury type, progress, and biometric feedback, for develop personal rehabilitation planning, as well as monitoring through smartphones or wearables to ensure patients perform exercises correctly, reducing re-injury risks. AI has also been widely applied in performance optimization, assisting development of tailored training programs, integrates data (sleep, nutrition, heart rate) to create dynamic training regimens that balance intensity and recovery, as well as providing real-time feedback through wearables for instant insights on hydration, fatigue, or technique during workouts. Recently, sports medicine practitioners have making use of AI in concussion management, for diagnosis, monitoring, analyzes balance, cognitive tests, or speech patterns to detect concussions and track recovery, informing return-to-play decisions. In elite sports talent identification, AI has been deployed for predictive scouting, using machine learning to evaluate physical/technical metrics (e.g., speed, coordination) in youth athletes to predict sport-specific potential. AI can provide personalized nutrition & recovery recommendations by suggesting diets, hydration, or recovery strategies based on metabolic data, activity levels, and goals. Many rehabilitation personnel have making use of AI to guide individual rehabilitation remotely (Virtual Therapy & Telerehabilitation), via virtual assistants or apps to guide patients through exercises remotely, using cameras to correct form and track progress. Last but not the least, many sports medicine practitioners are using AI for big data analysis in researches, to identify trends in injury epidemiology or treatment efficacy by aggregating data from global studies and electronic health records.
Similar to AI application in many disciplines, with rapidly growing applications of AI, there are also challenges & considerations in engaging AI in sports & exercise medicine. Ethical & privacy issues in ensuring secure handling of sensitive athlete data, how to integrate Human-AI collaboration with AI supplementing & enhancing clinician practice, rather than replaces clinician expertise, and how to avoid bias & inaccuracy to avoid skewed predictions, are some of the concerns that we have to handle. Nevertheless, I do strongly believe that AI's integration into sports & exercise medicine is driving precision, personalization, and proactive care, ultimately improving outcomes for athletes and patients alike.

The adoption of artificial intelligence (AI) into academic writing offers transformative opportunities but also poses complex challenges, requiring a careful balance between innovation and upholding of ethical standards. This speech will examine the dual-edged influence of AI tools, such as ChatGPT, Deep Seek, Grammarly, and DeepL Translate, in scholarly communication. We also evaluate existing policies from top-notched journals such as NEJM, Lancet and JAMA that seek to regulate their use for the authors.


AI provides extremely accelerated efficiencies which allows academics to draft manuscripts, refine language and phrasing, perform complicated data analysis, and retrieve literature at unprecedented speeds. For researchers where their native language is not English, AI plays a pivotal role in linguistics, enhancing more structured presentation and clarity. In addition, AI-driven platforms may inform identification of research gaps and generation of research hypotheses – which could help foster creativity. Nevertheless, these competitive edges are tempered by substantial concerns regarding practical and ethical considerations. Editors of Journals encounter issues related to authorship, originality, and accountability when the content of an article is completely dominated by AI: should AI be considered as a co-author? Are there any measures to avoid biases or plagiarism that may breach intellectual property rights? Will the increasingly common use of AI be downplaying critical thinking and compromising the human voice crucial to scholarly discussion? 


Many reputable journals have set up stringent policies that mandate transparency, necessitating disclosure of AI usage, and do not allow AI from being listed as an author. These regulations aim to maintain accountability, ensuring researchers should remain responsible for integrity and probity. However, journal guidelines on AI use are inconsistent and may cause confusion. No software seems to be able to confidently ascertain a case of AI use despite the availability of AI similarity index.


This speech calls for a collaborative framework where investigators should leverage the potential of AI while adhering to ethical standards, and all stakeholders including the press and publishers could standardize regulations to strengthen trust. We further argue that AI should be used as a tool but not a dominant platform to compose manuscripts and proposals. Measures should be taken to encourage human intellect in writing articles to achieve scholarly excellence.