While artificial intelligence is demonstrating remarkable potential to improve healthcare outcomes—from tuberculosis detection in remote Pacific islands to accelerating drug discovery—experts caution that without deliberate action, the technology threatens to widen the gap between well-resourced and underserved communities.
The warning comes as AI tools are being rapidly integrated into healthcare systems worldwide, raising urgent questions about equitable access, environmental impact, and the need for comprehensive regulation.
Transformative Applications
AI’s impact on global health has produced striking examples of innovation. In Chuuk state, Micronesia, an AI-assisted portable X-ray machine has transformed tuberculosis detection in remote islands where traditional diagnosis was impossible due to lack of electricity, medical specialists, and transportable equipment.
The technology enables health workers to test and immediately diagnose tuberculosis in seconds. Nearly everyone identified with active tuberculosis in the region has since begun treatment, addressing what had been a silent health crisis for decades.
Globally, AI is cutting clinical trial times in half, aiding antibiotic discovery, and improving disease outbreak predictions, including those linked to climate change. At a time when health systems—particularly in developing countries—face shrinking budgets alongside growing demand, AI presents opportunities to improve outcomes and drive health equity.
The Digital Divide Challenge
However, the same capabilities that make AI transformative also create risks of disruption and harm. The global digital divide represents the most significant barrier to equitable AI adoption in healthcare.
Countries and communities with constrained resources face unreliable or non-existent connectivity, inconsistent electricity and water supply, and prohibitive hardware costs. These infrastructure gaps undermine systematic, large-scale AI adoption that could improve health systems.
Without efforts to address infrastructure and affordability, experts warn that AI will widen existing health inequalities rather than narrow them. Underserved communities risk further marginalization, creating what critics describe as a two-tier system where innovation bypasses those with greatest need.
Mental Health Concerns
Even where AI is accessible, its deployment may generate unintended harms. AI chatbots capable of producing disinformation and misinformation pose particular risks for people with poor access to health professionals and therapy.
Some research indicates AI chatbots can positively affect psychological distress among young people. However, emerging evidence suggests they may exacerbate delusions in vulnerable individuals. This uncertain landscape demands rigorous research into optimal deployment of AI tools affecting psychological wellbeing, according to health policy analysts.
Environmental Impact
AI could also intensify disparities through its contribution to climate change, a threat disproportionately affecting underserved communities. Training and running large AI models requires enormous energy and water resources, increasing emissions and straining local ecosystems.
Experts argue that healthcare AI must be designed for efficiency, powered by renewable energy, and assessed for full environmental cost. The sector cannot improve health outcomes today at the expense of planetary systems that determine future health, according to this analysis.
Regulatory Developments
Regulation ensuring responsible AI use in healthcare is overdue, given the technology’s rapid embedding in global health systems. Effective regulation must include several components, experts suggest:
Incentivizing investments in digital infrastructure and reliable electricity to ensure equitable access
Establishing ethical safeguards for AI-enabled mental health tools to prevent unintended harms
Implementing a “green AI” framework driving investment toward energy-efficient models
Mandating accountability through “ethical traceability,” requiring disclosure of datasets, algorithms, and environmental impacts
Ensuring developing countries retain ownership and benefit from their health and environmental data
Promising developments include the United Kingdom’s recent participation in the HealthAI Global Regulatory Network, which enables regulators to share safety warnings, monitor real-world performance, and shape international standards collaboratively.
In the Asia-Pacific region, South Korea became the first country to adopt comprehensive AI legislation including safeguards for “high impact” AI systems in sectors such as healthcare. Multiple African countries, including Ghana, Tanzania, and Uganda, have similar legislation under development.
Narrowing Window
As AI continues developing rapidly and delivering public health benefits worldwide, experts emphasize that efforts to address its challenges must keep pace.
Without deliberate action, the benefits of AI risk being undermined by the harms it may introduce. The window for shaping AI’s integration into global health systems is narrowing, with decisions made now determining whether AI becomes a tool for health equity or a driver of deeper disparities, according to health policy observers.
