From the unknown to informed confidence: youth reflections on ethical AI, governance and the future of health

News & insights
April 20, 2026
Yifan Zhou
Erza Selmani
Sohayla Eldeeb
By -

In September 2025, DTH-Lab youth representatives Yifan Zhou, Erza Selmani and Sohayla Eldeeb were invited to join Diplo’s AI Apprenticeship: Bridging Generations in AI & Digital Diplomacy. As they wrap up their nine-week journey, they reflect on the key takeaways - including the urgent need for ethics and governance to be embedded into AI development to prevent online harm and safeguard data..  

The AI Apprenticeship is an online 9-week course initially designed for diplomats and experienced officials. As Regional Youth Champion and Research Fellow alumni, we had the opportunity to represent DTH-Lab in this intergenerational and cross-disciplinary cohort. Our goal was to learn from DiploFoundation’s expertise in building inclusive governance and policymaking, and connect this back to DTH-Lab’s goals of (1) empowering young people in digital, health and civic spaces to become active changemakers and co-create health futures; and, (2) fostering intergenerational exchange to reflect the needs and wants of young people in health policies.

Before starting the programme, we were worried about how technical the course would be. To our surprise, the course went beyond technical aspects and combined theory and hands-on experimentation to demystify AI. We learned not just what AI is, but examined in depth how Large Language Models (LLMs) work, learn, ‘reason’ and communicate. Interacting in a diverse cohort, we also learned how our peers in diplomacy saw AI through a governance and policymaking lens. 

Being part of the AI Apprenticeship Program at DiploFoundation added greater dimension and depth to our understanding of AI, including theoretical, governance and philosophical aspects, which enabled us to be more critical of AI.

Quality of data and intellectual property should be key foundations of AI

Quality of data and intellectual property should be key foundations of AI

Many may be familiar with the phrase “garbage in; garbage out” - when the input data is of low quality, the output is also of low quality. The same holds true with biased data; when the input data is biased, the output is also biased. The first takeaway was to minimize or eliminate bias. However, there is no universal singular truth on many matters. Who is to decide whether an action is appropriate or inappropriate? The cultural, ethnic, educational,  and/or personal experiences behind the data that AI is trained on make a difference on the response. Therefore, the second lesson is that instead of aiming to eliminate all bias, it would be much more constructive to attribute information to individuals or entities, so we can take these factors into account when interpreting the responses. With that in mind, our agents were designed to cite sources, flag uncertainty and surface regulatory or ethical considerations alongside scientific interpretation. 

Apart from the quality of data, quality data and, therefore, intellectual property are more valuable than ever in this AI era. Our work, thoughts and opinions can easily be stolen by AI, and become the property of large corporations without any credit. Before engaging with AI, it is crucial to have proper regulation on information usage and sharing to safeguard our knowledge from ‘theft’. Relating it back to healthcare, there is a deep hub of knowledge shared between indigenous populations around the world. Before plugging in their knowledge blindly into AI models, there needs to be protection of their knowledge and ownership.

AI: assistant rather than authority

AI: assistant rather than authority

Will AI replace healthcare professionals? The short answer is no. AI is nowhere near replacing them. They are mathematical models that compute probability, rather than having a true understanding of knowledge, context, and reasoning like we humans do. Therefore, we should see AI as assistive, rather than authoritative.

AI is known to hallucinate information that is inaccurate or nonexistent. Interestingly, it is correlated with AI’s creativity. When interpreting AI’s responses in healthcare contexts, we should be mindful of its potential biases and hallucinations.

On the other hand, we can use AI to enhance our own thinking, for example, by applying Socratic inquiry to our interactions with chatbots. This helps us reflect on our own assumptions and thought processes. It was intriguing to see how philosophy, an ancient study, connects so closely with AI, the novelty of this era.

Ethical AI development is critical to preventing online harm

Ethical AI development is critical to preventing online harm

Prior to the course, many of our fears surrounding AI arose from the unknown: how models are trained, where bias enters and what algorithms can and cannot do. Now, we are concerned because every technical and ethical decision in AI development can make the difference between real benefit and real harm. Without proper governance and safeguards, the impact on young people’s online safety and well-being can be detrimental.

Through conceptualizing theory, this course gave shape and form to the ‘unknown’, and our fear is replaced with informed confidence. In a digital age, having more than a superficial understanding and the ability to critically appraise AI is key to taking agency in shaping how AI transforms our systems.

Find out more about DTH-Lab’s work on digital and data governance https://www.dthlab.org/digital-and-data-governance

Click on the links below to read the reports:
No items found.
No items found.
References

Yifan Zhou, PharmD, ACPR, is a pharmacist in Canada, where she practices both in the industry and hospital. As the immediate past Chairperson of External Relations at International Pharmaceutical Students’ Federation (IPSF), she managed over 20 partner and sponsors relations. She has also represented the Federation at more than 10 high level international meetings, where she delivered statements, interviewed, and advocated for public health, climate change, digital health, and meaningful youth engagement.Yifan is currently the President-Elect of IPSF, where she hopes to continue advocating for 500,000 pharmacy students and recent graduates around the world.She was the past Executive Co-Chair of IMAGINE Clinic, a student-run clinic that serves marginalized community in Toronto. During the pandemic, she led a team of 30 executives to transform the walk-in clinic into a virtual clinic to continue providing care. The clinic also offered physical space, computer and internet access to bridge the gaps in access for its patients. Yifan also developed an automated flu form for a pharmacy technology company, which served over 700k Canadians during the pandemic.

Erza is a researcher and consultant in the health and social domain. She studied digital health systems in the UK and is currently doing her PhD studies in Germany with a focus on digital and robotic interventions in health. She is engaged in several international projects in health research and development. Her research interests revolve around the health and social science fields, digital technologies that improve health/social well-being, social prescribing and user-centred design. She has co-authored several publications in peer review journals.

Sohayla Eldeeb is an environmental health advocate focusing on the intersection of health and sustainability, specifically how digital health plays a role in net zero. She graduated from Stanford with a bachelor's in biology and a master's in Community Health and Preventive Research. In 2018, Sohayla was one of the early team members of Zero Hour, the first youth-led climate march on Capitol Hill in Washington, DC, and formed over 200 action chapters worldwide. Since then, she has pushed for reusable gowns and lower-emission food/procurement systems at Stanford. Sohayla has helped lead several initiatives revolutionizing the healthcare systems to net zero, including an innovative greenhouse gas emissions (GHG) calculator for telehealth and transitioning to healthier food systems. She has been named a Healthcare Sustainability Champion. She is extremely passionate about the future of health tech, specifically how it may be used to integrate more representative health measures of social and environmental determinants. Sohayla has spoken at COP 27 about the human right to a healthy environment, the first-ever United Nations Youth Climate Summit, and served as a youth speaker at Africa's CDC Health Conference.

You may also like

Publication
May 6, 2026
DTH-Lab Annual Report 2025
Publication
April 29, 2026
Everyday Digital Technology Use and Youth Health: Scoping Review of Longitudinal Studies
Publication
April 27, 2026
Co-creation - A new standard in working with youth as partners in designing health futures
News & insights
April 20, 2026
From the unknown to informed confidence: youth reflections on ethical AI, governance and the future of health
News & insights
April 13, 2026
From tobacco to TikTok: what public health litigation history tells us about holding social media accountable
News & insights
April 6, 2026
Making young people visible: what national digital health strategies reveal and how a new framework can respond
Multimedia
March 31, 2026
Session 2: Governing Digital Health Systems to Promote Solidarity, Equity and Youth Well-Being - Research Fellows dissemination series 2026
News & insights
March 30, 2026
DTH-Lab Fellowship Programme applications open now
News & insights
March 27, 2026
Why Nigeria must address the digital determinants of health
Multimedia
March 24, 2026
Session 1: Designing Healthy Digital Futures for Young People - Research Fellows dissemination series 2026
News & insights
March 12, 2026
Sharing our latest Research Fellows’ publications
News & insights
February 27, 2026
Scholarship Opportunity for MyHealthFutures Youth Network Members
Publication
February 18, 2026
Harnessing data for smarter antimicrobial stewardship: A case study on data solidarity in action
Publication
February 17, 2026
Digital childhood profiles 2026
News & insights
February 13, 2026
From agency to accountability: aligning caring masculinity with digital regulation
Publication
February 11, 2026
Measuring meaningful youth engagement and impact
Publication
February 11, 2026
Engaging youth in shaping health futures
Publication
February 4, 2026
Healthy digital diets: A review of existing guidelines for young people’s technology use
Multimedia
January 26, 2026
Prince Mahidol Award Conference 2026 Posters
News & insights
January 22, 2026
DTH-Lab announces strategic partnerships with IFMSA and IPSF to strengthen the future health workforce