Advancements in human genetics and genomics have been continually impacting basic human rights and fundamental freedoms, including privacy, confidentiality, non-discrimination, and human dignity. The increased application of artificial intelligence (AI) to human genetics and genomics has intensified these concerns in recent years. AI technology that poses risks to human existence and which continually challenges (and often surpasses) the cognitive abilities of human beings raises a debate over the coexistence and mutual respect of AI and humans.

The regulatory debate on AI in the context of human genetics and genomics involves two continuums — risks and innovation — leading to contentious discussions over global, local, and self-regulation. Given the potential benefits of AI, companies and even countries are investing in and promoting the development of AI. It is encouraging that companies are volunteering to develop ethical frameworks for AI regulation; however, they often miss the core values human rights protect. It triggers an inquiry into why global governance of AI to human genetics and genomics is crucial, and how the human rights framework can provide effective global governance.

Human genomics involves the increased application of data-driven science, where the application of AI is manifold, ranging from diagnosis and screening to drug development and treatment. It holds great promise in healthcare by adding precision, speed, and optimisation to genetic diagnosis, precision medicine, and precision health. It can help promote the next wave of medical and healthcare immensely.

However, it can also deepen and enlarge human rights concerns, such as data privacy and biases in medical algorithms and decision-making. Diversity in genetic datasets is important for creating algorithms used to train machine learning models. The lack of diversity in global genetic datasets and the exclusion of marginalised populations in national genetic datasets may lead to discrimination and inequity in genetic research and genomic medicine. Thus, research findings and treatment using these datasets may be inaccurate and less useful for the excluded populations. It may also create a new subclass, privileged with the genetically advanced treatment, leaving others from its loop and deepening health disparities.

Furthermore, the black-box problem (that is, the difficulty in comprehending how a complex system arrives at a decision) complicates the explainability of AI-driven decision-making in the field of human genetics and genomics. Although existing ethical, legal, and policy frameworks address developments in the domain of human genetics and genomics, notably the Universal Declaration on the Human Genome and Human Rights (1997) and the Declaration of 2002, followed by regional human rights instruments and specific legislations of various countries. However, new and unprecedented challenges arising from the AI-human genetic and genomic interface warrant a specific approach. At the intersection of AI and genetics, robust data anonymisation is extremely difficult, and AI’s surveillance, data tracking, and pattern recognition capabilities exacerbate human rights risks.

The relevance of a human rights gaze into the regulation and governance of AI-human genetics/genomics is pertinent on several counts. The universality of human rights appears to be compatible with the universal (cross-border) nature and global impact of AI and genetics. The human rights approach takes a human-centric perspective, whereas technology-driven policies guide ethical frameworks that are developed and implemented by companies with profit as the primary focus. While ethical frameworks are open to varied interpretation, human rights are more established and uniform in their application. Finally, contrary to human rights, which are enforceable and actionable, ethical frameworks appear weak due to their reliance on self-regulation and principles that can be ambiguous and unclear.

The right to science, which has been recognised as a cultural right in international human rights instruments, could help regulate AI and genetics equitably, as it ensures the benefits of science are available to all. In the genetic and genomic research, data stewardship and data security also require inspiration from the HR framework. Scientists need to be more committed to respecting the human rights of individuals and groups who participate in their experiments, whether willingly or not.

Given the serious impact of AI on humanity (such as autonomous weapon systems, and genetic discrimination), human rights norms become essential and should therefore support the global governance of AI. AI complements humans; therefore, a debate about their harmonious and respectful coexistence makes sense. Human rights-centric AI regulation has a blend of coexistence, managing vast genomic datasets, harmonising and standardising this information, and employing machine learning techniques to clean missing data and sort human genomic data.