In March and April 2025, the Forum for Medical Ethics Society (FMES), the Indian Journal of Medical Ethics (IJME), the Health, Ethics and Law (HEaL) Institute, and the Christian Medical College Vellore (CMCV) co-organised the 10th National Bioethics Conference on “Artificial Intelligence in Health Care: Ethics, Law and Human Rights Matters”. This conference report summarises the key discussions and ethical concerns raised during the presentations.
Keywords: artificial intelligence (AI), ethics, governance, health care, human rights
Since the inception of the National Bioethics Conference (NBC) in November 2005, this biennial gathering has brought together individuals and organisations to discuss bioethics concerns in India and elsewhere, focusing on issues related to the governance of healthcare, research ethics, medical technologies and more. Past NBCs were held in Mumbai, Bengaluru, Delhi, Hyderabad and Pune before being moved to the virtual mode during the Covid-19 pandemic in 2020. The ninth NBC was held in hybrid mode, in Chennai and online [1].
The theme of the 10th NBC was: “Artificial Intelligence in Health Care: Ethics, Law and Human Rights Matters.” The conference aimed to start a discussion on the potential impact of Artificial Intelligence (AI) technologies on healthcare, issues of ethics, human rights and regulation, and existing guidelines and regulatory bodies of AI-health governance. The conference, held online over March-April 2025, was a collaboration of the Indian Journal for Medical Ethics (IJME), Health, Ethics and Law (HEaL) Institute, Forum for Medical Ethics Society (FMES) and the Christian Medical College (CMC) Vellore. The speakers included experts in bioethics, medicine, public health, health policy and governance, digital and global health, and more [2]. The conference was attended by 198 participants.
According to a 2024 World Health Organization (WHO) report, AI refers to the “capability of algorithms integrated into systems and tools to learn from data so that they can perform automated tasks without explicit programming of every step by a human” [3]. With such a capacity, AI in healthcare is expected to predict outcomes, identify patterns, and extract clinically relevant actionable information while leveraging advanced statistical algorithms. With advanced AI technologies like Generative AI (GenAI), which creates new content based on the data it is fed, and Large Language Models (LLMs) which use techniques like neural networks (computer architecture modelled on the human brain to mimic human-like learning in machines) for processing human language, the scope of AI in healthcare is rapidly evolving.
AI uses in the healthcare sector include diagnosis and clinical care, clerical tasks, research and development, and medical education [1]. The potential impact of these uses has stimulated debate on the legal and ethical concerns.
The opening session began with Dr Arul Dhas T, Consultant in the Department of Bioethics at CMC Vellore, who emphasised the need to align AI with values such as health and justice. Dr Sanjay Nagral, chairperson of FMES, reflected on FMES and IJME’s backgrounds and evolution. He noted that IJME aimed to maintain a commitment to open-access scholarship and grassroots ethics. Dr Sunita Sheel, Director of HEaL Institute and a working editor of IJME, positioned Indian bioethics as a people’s movement and highlighted the role of NGOs, which combine academic work in health and activism. Sayantan Datta and Lubna Duggal, both working editors of IJME, previewed the themes for each day and highlighted the growing gap between the advances in AI technology and the ethical safeguards being developed for its use. Duggal stated that this disconnect was central to the conference’s theme. During this session, the FMES-IJME Ethics Awards were presented to Dr Lopa Mehta, former Head of Anatomy at GS Medical College, Mumbai, and Dr Anant Phadke, founding member of All-Indian Drug Action Network (AIDAN) and long-term volunteer of the Peoples’ Health and Sciences Movement. In her acceptance speech, Dr Mehta reflected on the implications of AI’s inability to understand suffering, a requirement central to medicine. Dr Phadke’s comments focused on economic justice, as he called for healthcare to remain a public good in an increasingly corporatised digital world.
Plenary 1: Health technology assessment and AI: Past and presentThe first plenary was chaired by Dr Mala Ramanathan, Professor at the Achutha Menon Centre for Health Science Studies, Thiruvananthapuram and working editor of IJME, and Dr Sandra Albert, Director of the Indian Institute of Public Health, Shillong.
Dr Indranil, Professor at the School of Government and Public Policy, OP Jindal Global University, Sonipat, opened the session, asserting that AI is neither “artificial” nor “intelligent,” but a tool that often reduces human complexity to data. He warned against “elite capture,” a process where policies, resources, or technologies benefit powerful groups while excluding the marginalised. In the context of AI, this means that the development and control of algorithms always lie in the hands of large corporations. He connected this to the rise of “information capitalism,” where personal data becomes monetised. Dr Shankar Prinja, Professor of Health Economics at the Post Graduate Institute of Medical Education and Research, Chandigarh, discussed Health Technology Assessment (HTA) of AI tools in healthcare. HTA is used to evaluate the clinical, economic, and social value of medical technologies with methods such as clinical trials, cost-effectiveness analysis, and ethical impact assessments. Prinja pointed out the challenges of assessing constantly evolving AI when assessment is a slower process. He stressed that AI tools need constant oversight and not just one-time approval by regulatory frameworks. Dr Calvin Wai-Loon Ho, Associate Professor at Monash University, Australia, reported on the UK’s NHS-DeepMind, an app which used AI to diagnose eye diseases at an early stage. DeepMind is a “learning healthcare system” which helps in increasing the accuracy of AI. Researchers continuously feed anonymised clinical data into the system, creating a feedback loop where doctors confirm or correct its diagnoses.
Session 1: AI applications and psychiatry: Ethics, governance, and entrepreneurshipsThe first session, chaired by Dr Sunita Simon Kurpad, Professor of Psychiatry and Ethics at St. John’s Medical College and Hospital, Bengaluru, and Dr Sunita Sheel shifted the focus to mental health.
Dr Suresh Bada Math, Professor of Psychiatry at NIMHANS, Bengaluru, laid the foundations for this discussion by pointing to AI’s transformative potential in psychiatry through improved diagnostics and therapy outcomes. However, he cautioned against overreliance on AI tools which have so far not been able to communicate the empathy that is essential to therapy. He also emphasised the need for strict adherence to Indian Council of Medical Research (ICMR) guidelines [4] on autonomy, privacy, and human oversight. Dr Smriti Joshi, a psychologist and board member at Wysa, a platform which uses AI-powered tools to offer mental health support, offered a practical example of ethical AI deployment. She described Wysa’s use of structured Cognitive Behavioural Therapy (CBT) scripts. Wysa integrates CBT techniques into its AI-enhanced conversations, thus helping users recognise negative thought patterns. Platforms such as Wysa can make mental health support stigma-free and reach those who may be hesitant to seek traditional therapy. However, she admitted that corporate pressures and global demands often pose challenges. Dr Dilip Jeste, Director of the Global Research Network on Social Determinants of Mental Health and Exposomics, spoke on AI’s lack of “wisdom.” He described wisdom as rooted in compassion and empathy. Drawing on neurobiological studies, he suggested that wisdom extends beyond cognitive abilities and is closely linked to brain regions responsible for emotional regulation. He pointed to the prevalence of loneliness due to factors like the rise of social media and referred to research suggesting that people who experience greater loneliness tend to have lower levels of wisdom. A concern was raised about AI: can machines lacking lived experience and emotional depth ever be wise? If modern social conditions are hindering human wisdom, then building AI that reflects empathy and ethical judgement becomes more difficult. A wise AI system would need to understand human emotions, learn from its mistakes, and integrate diverse viewpoints. Ultimately, he emphasised the need for multidisciplinary collaborations between technologists, clinicians, ethicists, and mental health professionals.
The plenary was chaired by Dr Nandini K Kumar, bioethicist and President, Forum for Ethics Review Committees in India, and Dr Joy John Mammen, pathologist and biomedical informatics expert at CMC Vellore.
Dr Rohit Malpani, an independent consultant working on AI ethics, access to medicines and global health policy, and Dr Andreas Reis, co-lead of the Health Ethics & Governance Unit at WHO focused on WHO’s 2021 guidelines on AI in healthcare [5]. The guidelines identify six key ethical principles: i) protecting human autonomy by ensuring that humans control AI; ii) promoting human well-being and safety; iii) ensuring transparency and explainability; iv) fostering responsibility and accountability from AI developers; v) ensuring inclusiveness and equity to minimise bias; and vi) promoting responsiveness and sustainability. The WHO examined the potential benefits of LLMs taking over clerical tasks, while keeping in mind risks such as “hallucinations”, or the generation of incorrect information by the AI tool. Dr Siby K George, Professor of Philosophy at the Indian Institute of Technology Bombay, stated that while humans originally created and controlled technology, they are now being influenced by it. He pointed out that AI’s use of patient data raises significant ethical concerns since its decision-making process is not understood. George warned that AI’s biases could deepen inequalities if treated solely as technical issues. We cannot ignore the ethical questions about how the decisions are made and what populations could be affected. He stressed the need to critically assess AI tools based on whether they prioritise public health or profit.
Session 2: Health digitisation in India: Voice from the grassroots and policy spacesThe second session was chaired by Dr Sylvia Karpagam, public health physician and working editor of IJME, and Dr Hammad Durrani, a global and digital health expert currently serving as Technical Advisor at Nutrition International, Canada.
Dr Sreerupa, Research Fellow and Program Lead at the Institute of Social Studies Trust, New Delhi, focused on how the lives of frontline healthcare workers — Accredited Social Health Activists (ASHA workers) — are impacted by the push for digital technology. Since AI relies on data for training, community health workers are being burdened with the additional responsibility of collecting and managing data. She and her team conducted studies on ASHA workers’ experiences across multiple states which revealed different experiences with digitalisation (using digital tools to improve processes and services, for example, streamlining software); the challenges faced included the pressure on ASHAs to rapidly transform into “smart ASHAs”, or digitally skilled workers, but without recognition or extra pay. Another consequence was the exclusion of older or less literate workers. The recommendations to improve the process included training techniques that take into account gender and social contexts and involving workers in the designing of the digital systems they use. Dr Narendra Gupta, senior community health physician, pointed out that the current model of healthcare digitisation (storing analogue data in an electronic format, for example, scanning documents) in India has significant challenges at the grassroots level because of the difference in understanding digital tools in rural versus urban areas. Digital systems often prioritise data collection over its meaningful use. Hence, Gupta suggested digitisation be implemented in populations that demand it, where people need and use it. He also stated that effective digitisation requires decentralisation, empowering local healthcare workers and communities. Dr Usha Ramanathan, legal researcher and human rights expert working on India’s national ID project, offered a different perspective on digitisation. She focused on a fundamental question: Why collect so much data if it is not being used meaningfully? Over time it has become clear that the purpose of digitisation was to serve private rather than public interests. Ramanathan presented the timeline of evolving legislation from 2005 to 2025, highlighting changes in rules on the usage of data by private and public sectors. An example brought up was how insurance companies and pharmaceutical firms use the health data that is linked to digital IDs to target specific demographics.
The plenary on day 3 was chaired by Vivek Divan, coordinator at Centre for Health Equity, Law & Policy (C-HELP), and Jai Ganesh Udayasankaran, Executive Director of Asia e-Health Information Network (AeHIN).
Anita Gurumurthy, a founding member and Executive Director of IT for Change, a Bengaluru-based non-profit, began by highlighting India’s underinvestment in public health, questioning whether its fragile infrastructure can meaningfully integrate cutting-edge technology without worsening inequalities. She stressed that the privatisation of health information, accelerated by initiatives like the Ayushman Bharat Digital Mission (ABDM) [6], concentrates power among corporations and policymakers. While ABDM seeks to centralise India’s health data for improved public health and AI-driven innovation, Gurumurthy argued it lacks safeguards for autonomy and consent, criticising a “No Fee, No Data, No Service” model that ties care access to mandatory data sharing, often through Aadhaar linkage. Dr Barry Solaiman, Associate Dean for Academic Affairs at Hamad Bin Khalifa Law University, Qatar, discussed the legal and regulatory challenges in governing health AI. He emphasised that the current guidelines lack enforceability, giving rise to legal concerns such as data privacy breaches, algorithmic bias, medical liability, and the challenge of obtaining true informed consent when patients are often unaware of AI’s involvement in their care. Comparing regulatory landscapes, he observed that cultural and religious contexts complicate the efforts of standardised global AI regulations. He concluded that futureproofing requires shifting from product-specific regulation to governance of the entire AI lifecycle.
Session 3: AI and marginality: Users’ perspectives from transgender persons and people with disabilitiesThis session moved the conversation from policy-level discussions to the lived realities of marginalised groups in times of AI boom. It was chaired by Sayantan Datta, and Shampa Sengupta, founder of Sruti Disability Rights Centre.
Mridul D, a freelance full-stack tech professional, noted that AI technologies often reproduce biases, risking misdiagnoses for transgender persons and people with disabilities, particularly because of non-representative datasets. He outlined the ways in which AI is embedded into health ecosystems, including symptom checkers, chatbots, assistive technologies, and insurance systems. Furthermore, AI-driven surveillance and data scraping — importing information from a website into a local file — violate the privacy of transgender individuals, exposing their identities without consent, thus compounding vulnerabilities. Mukul Pandya, an Associate Fellow at Oxford University’s Saïd Business School, offered a first-person narrative about rebuilding life after a stroke affected his writing and editing capabilities, skills central to his professional identity. He described how technologies such as WhatsApp’s voice messaging and transcription tools like Otter.ai enabled him to gradually regain agency. He proposed that when used thoughtfully, AI can become a vital tool for rehabilitation and the reclaiming of selfhood, especially for people grappling with severe disabilities. Abhiti (Ell) Gupta (they/them), an independent consultant working at the intersections of gender, sexuality, health and law, acknowledged that AI can promote independence and participation among persons with disabilities as well as trans people. However, systemic barriers remain deeply entrenched. The limited smartphone ownership in these communities, the lack of accessible infrastructure and affordability, and underdeveloped welfare schemes all severely limit the reach and effectiveness of assistive technologies. Gupta called for welfare and healthcare frameworks to move beyond token inclusion to meaningful accessibility; they highlighted the need for awareness of the discriminatory practices built into supposedly neutral technological systems. Importantly, the goal should not be to “mainstream the margins” but to challenge prevailing notions of “normalcy” that continue to exclude diverse bodies and identities.
The last session of the conference shifted focus from AI in healthcare to its usage in academia and publishing. The session was chaired by Lubna Duggal and Sayantan Datta.
Neha Mishra, Assistant Professor of Practice at the Centre for Writing and Pedagogy, Krea University, argued that we must ask if AI should be integrated into classrooms at all. Mishra raised concerns about genAI’s impact on academic rigour by emphasising the risks of inaccuracies, plagiarism, and the unaccountability behind knowledge production. Using the feminist critique of unchecked technological progress, she argued that technology contributes to the systems of control following the patriarchal systems upon which it was built. She argued that the introduction of AI into academia should not be treated as inevitable; we must resist technological progress when it compromises academic standards. Dr Sunaina Singh, academic trainer and scientific editor, focused on how AI could be used responsibly. Singh argued that AI can take over mundane tasks within research and writing, such as citing, quality checks and data extraction, so that researchers can carry out higher-order thinking with more rigour. She maintained that AI can enhance efficiency, but it should only augment, and not replace, human judgement. Dr Piyali Mitra, Deputy Editor of the Asian Bioethics Review, argued that LLMs, which lack consciousness and conscientiousness cannot be authors, only tools. They are disqualified from being legal entities, and cannot hold copyrights under Indian law. Lastly, Chris Zielinski, President of the World Association of Medical Editors, spoke on the association’s approach to regulating AI usage in publications. He argued for the need to code ethical considerations into AI and treat the issue at its core instead of only framing regulations around AI applications.
Dr Alvin B Marcelo, head of the Asia eHealth Information Network, discussed the integration of AI into medical education and research. He highlighted the benefits of personalised learning, simulation-based training, and access to knowledge, while also raising concerns about ethics and over-reliance on AI. While AI can enhance a person’s understanding of topics, it can also lead to misinterpretation, especially in sensitive environments like doctor-patient interactions, where tone is important. Dr Chi Yeung Eric Ip, bioethicist and Professor of Law at the University of Hong Kong, focused on the ethical foundations of digital planetary health. He called for immediate action on human-induced climate crises, discussing platforms such as the World Environment Situation Room and the Global Environment Monitoring System, which gather data for global decision-making. He also spoke on the European Commission’s Destination Earth project for predictive climate modelling.
The conference highlighted the concern that while AI promises advances in efficiency, diagnostics, and innovation, its deployment is often shaped by market forces rather than public health. For example, Anita Gurumurthy emphasised how the ABDM, despite being framed as a public initiative, facilitates data extraction and privatisation of health data into a transactional commodity. Dr Rohit Malpani and Dr Andreas Reis expanded this critique to AI in pharmaceutical R&D, where innovations in health are driven by their commercial potential, deepening global inequities. Dr Siby K George foregrounded the philosophical stakes: when human creators lose control over the technologies they design, autonomy and moral accountability are undermined. This shows that AI systems, if guided primarily by commercial logic, undermine the ethical foundations of healthcare. Together, these reflections from the conference caution against viewing AI as neutral. Its governance must be shaped by public interest, justice and social responsibility.
The artificiality of AI and the lack of accountabilityAnother theme in the conference was that the non-human nature of AI made it impossible to hold it accountable. This led to apprehensions about its applications in the medical field, where accountability is a core principle. Most speakers maintained that despite undergoing continuous rigorous testing, AI cannot be trusted with the ultimate decision-making responsibility. Both Dr Indranil and Dr Piyali Mitra’s arguments about AI not being qualified bring up a core question: will we ever reach a point where AI can be more than an information processing unit, thus being capable of accountability? Dr Dilip Jeste’s vision of “wise robots” suggested this may be possible. However, Neha Mishra’s presentation showed us that even if it were possible, it might not be desirable. Speakers reporting the experiences of marginalised groups showed that without active and rigorous supervision, the biases embedded in AI can create new systems of oppression. Thus, as the use of AI becomes more widespread, we must remain alert to its perils. We must remember that AI is ultimately a machine, and thus can only contribute in terms of information processing. It cannot replace human judgement in areas of decision making.
The 10th National Bioethics Conference revealed the challenges of integrating artificial intelligence into healthcare and academia, resisting one-dimensional views of AI as wholly progressive or dangerous. Speakers called for caution, stressing that the adoption of AI must not override principles of integrity, transparency, privacy, confidentiality, and consent. While the conference raised important criticisms of private-sector involvement in healthcare data systems, it appeared to lack dialogue with the developers, corporations, and regulators actively shaping AI deployment. The insights shared at the NBC collectively challenged the assumption that AI integration is inevitable or neutral. They also demonstrated that the ethical implications of AI cannot be addressed without first acknowledging social and political concerns. The developments in health AI must be concerned with more than just technical innovation.
Authors: Ananya KV (Ananya_venkatesh.sias23@krea.ac.in), Niharika Banerjee (Niharika_banerjee.sias23@krea.ac.in), Srestha Majumdar (Srestha_majumdar.sias23@krea.ac.in), Vyahruti Prakya (corresponding author — Vyahruti_prakya.sias23@krea.ac.in), Undergraduate Student, School of Interwoven Arts and Sciences, Krea University, 5655, Central Expressway, Sri City, Andhra Pradesh, 517646 INDIA
Conflict of Interest: None declared Funding: None
Author Notes: The authors contributed equally to this manuscript.
Acknowledgments: The authors thank Sandhya Srinivasan and Sayantan Datta for their guidance and support in writing this manuscript.
To cite: Ananya KV, Banerjee N, Majumdar S, Prakya V. Artificial intelligence in health care: Ethics, law and human rights matters — an overview of the 10th NBC proceedings. Indian J Med Ethics. Published online first on November 4, 2025. DOI: 10.20529/IJME.2025.083
Submission received: August 10, 2025
Submission accepted: October 22, 2025
Manuscript Editor: Olinda Timms
Copyright and license
©Indian Journal of Medical Ethics 2025: Open Access and Distributed under the Creative Commons license (CC BY-NC-ND 4.0), which permits only noncommercial and non-modified sharing in any medium, provided the original author(s) and source are credited.
References