For a pregnant woman in a rural settlement in Kuje, Abuja, accessing specialist care may mean hours of travel, considerable expense, and uncertainty. For a child in a hard-to-reach community in Borno State, pneumonia may go undetected because a health worker lacks the necessary diagnostic tools. In a riverine community in Bayelsa State, a person with tuberculosis (TB) may face delays in diagnosis because radiology services are difficult to access. These scenarios reflect a fundamental challenge facing Nigeria's healthcare system: those who need timely medical attention most often face the greatest barriers to receiving it.
Again, consider a scenario in which a frontline health worker is equipped with an artificial intelligence (AI)-powered tool that facilitates early identification of complications, interpretation of medical images, or recognition of patients requiring urgent referrals. As African countries explore AI-enabled healthcare, a critical question emerges: is technology being adopted solely due to its availability, or because it demonstrably improves health outcomes? To be sure, the value of AI lies not in its sophistication, but in its ability to improve health outcomes, expand access to appropriate care and reduce inequalities.
AI offers opportunities to strengthen healthcare delivery, particularly where specialists, diagnostic equipment and other resources are limited. When properly validated and implemented, it can help health workers identify risks, interpret clinical information and make informed decisions. Tuberculosis screening illustrates this potential. AI-powered computer-aided detection (CAD) tools analyze chest X-rays to identify abnormalities suggestive of tuberculosis, helping health workers prioritize individuals for further testing. For eligible populations aged 15 years and older, these tools can improve screening and triage, particularly in settings with limited access to radiologists. However, screening must be followed by confirmatory testing, clinical assessment and timely linkage to treatment.
Nigeria offers encouraging evidence of this potential. A study involving 869 participants across 17 communities in the Federal Capital Territory evaluated an AI-enabled automated microscope for detecting Schistosoma haematobium eggs in urine, an indicator of urinary schistosomiasis. Compared with conventional microscopy, the tool demonstrated approximately 89 per cent sensitivity and 99 per cent specificity. Although promising, these findings require further validation in routine clinical settings. Affordability, reliability, quality assurance and scalability will determine whether such innovations can deliver sustained benefits. Put differently, success should be measured not only by AI's ability to detect disease, but by whether patients receive appropriate care as a result.
For many Nigerians, distance, transport costs, language barriers and shortages of health workers limit access to essential services. Voice-based AI could support frontline workers and improve access to reliable health information, provided these tools accommodate local languages, cultural contexts and limited internet connectivity. Community-informed design and validation are essential. AI should enhance, rather than replace, the trust between patients and health workers. Similarly, AI-enabled surveillance could help identify communities missed by routine immunization and improve polio response planning. However, data alone cannot vaccinate a child. Health workers, available vaccines and community trust remain indispensable.
AI's contribution must therefore be measured by how effectively it strengthens existing health systems and extends their reach. Nigeria's realities must guide the development and implementation of AI in healthcare. Unreliable electricity limited rural connectivity, fragmented health information systems, high costs, and shortages of technical expertise can undermine promising innovations. For a rural health facility, the practical questions are straightforward: Can the tool function during power interruptions? Can health workers use it with limited connectivity? Is technical support available when it fails? Will the service remain operational beyond the pilot phase? Implementation costs must extend beyond purchasing technology to include electricity, connectivity, system integration, staff training, maintenance, cybersecurity and regulatory compliance. Solutions dependent on short-term donor funding risk becoming unsustainable once external support ends.
Interoperability is equally important. Nigeria's digital health systems must securely exchange relevant information rather than operate in isolation. Integrated infrastructure can reduce duplication, improve continuity of care and make better use of limited resources. Local ownership is also critical. Nigeria's researchers, universities, health workers, regulators, government institutions and technology developers must actively shape these solutions. Building domestic expertise and innovation capacity will ensure that AI reflects national health priorities, languages and clinical realities. The benefits of AI must be balanced against its risks. Systems trained on datasets that inadequately represent African and Nigerian populations may produce unreliable results, reinforce existing health inequalities and exclude vulnerable groups, including women, persons with disabilities, ethnic minorities and underserved communities. Generative AI can also produce inaccurate information, while changes in disease patterns, clinical practices and data quality may affect performance over time. Consequently, responsible implementation requires independent evaluation, transparency, continuous monitoring and robust governance. Patient privacy and confidentiality must be protected, with clear accountability when AI-supported decisions contribute to harm. WHO guidance, including its 2021 guidance on ethics and governance of AI for health, 2023 guidance on regulatory considerations for AI in health, and 2024 guidance on large multi-modal models, provides direction for responsible development and use.
On their part, health workers must be equipped to understand AI's strengths and limitations, critically assess its recommendations and prioritize patients' best interests. AI should augment, not replace, human judgment. Technological recommendations must never override professional responsibility, and patients must remain at the centre of every healthcare decision.
To translate AI's promise into measurable health gains, Nigeria must focus on five priorities. First, prioritize health needs by targeting clearly defined health challenges instead of focusing on technological novelty. Second, scale what works. Ensure clinical validation, regulatory oversight, and clear pathways from screening to diagnosis, referral, and treatment. Third, strengthen digital infrastructure. Invest in reliable electricity, connectivity, secure health data and interoperable systems. Fourth, build local expertise. Empower Nigerian researchers and health professionals to develop, adapt and evaluate context-appropriate solutions. Fifth, ensure sustainable financing. Fund training, maintenance, technical support and long- term service delivery beyond pilot projects. Returning to the examples of the pregnant woman in Kuje, the child in Borno, and the individual awaiting a TB diagnosis in Bayelsa, success would entail earlier recognition of danger signs, reliable diagnostic services, timely referrals, and access to effective treatment, irrespective of geographic location or socioeconomic status.
AI can make Nigeria's health system more responsive, efficient and equitable, but algorithms alone cannot deliver healthcare. Progress depends on trained health workers, functioning facilities, reliable infrastructure, sound governance, sustainable investment and public trust. The central question is no longer simply whether AI can transform healthcare, but whether Nigeria can translate its capabilities into earlier diagnoses, timely treatment and better health outcomes, particularly for underserved populations. With its strong leadership, the Federal Ministry of Health and Social Welfare, working with relevant government institutions and partners, can establish and implement robust governance frameworks for AI in health. These should promote ethical use, patient safety, transparency, accountability, data protection, equity and meaningful human oversight. The World Health Organization will continue to support Nigeria in harnessing AI responsibly to strengthen healthcare delivery.