The Nigerian pharmaceutical landscape faces a critical juncture. As policymakers grapple with the dual challenge of ensuring drug safety while accelerating access to essential medicines, the National Agency for Food and Drug Administration and Control (NAFDAC) stands at the frontlines of a battle that directly impacts millions of lives. The stakes are extraordinarily high: research by Adigwe, Onavbavba, and Wilson published in the Integrated Pharmacy Research and Practice journal in 2022 found that pharmacists across Nigeria identified poor collaboration among regulatory agencies (89.94%), inadequate inspection (90.93%), and insufficient legislation (88.83%) as primary barriers to combating counterfeit medicines. The same study revealed that 39.6% of pharmacists had unintentionally purchased counterfeit medicines during their practice—a sobering statistic that underscores systemic vulnerabilities in the pharmaceutical supply chain.
Meanwhile, regulatory agencies in developed nations are pioneering artificial intelligence (AI) solutions that promise to revolutionize drug oversight. The question for Nigerian policymakers is no longer whether AI can transform pharmaceutical regulation, but rather how quickly and effectively NAFDAC can harness these technologies to protect public health.
NAFDAC’s Critical Role and Persistent Challenges
Established by Decree No. 15 of 1993 and subsequently amended, NAFDAC operates under the NAFDAC and Control Act Cap N1 Laws of the Federation of Nigeria (2004). Research by Agbo Friday Ojonugwa and Dr. Gwom Solomon Gwom published in the Wolverhampton Law Journal provides a comprehensive analysis of NAFDAC’s regulatory mandate. The agency is responsible for regulating and controlling the manufacture, importation, exportation, distribution, advertisement, sale, and use of food, drugs, cosmetics, medical devices, packaged water, chemicals, and detergents—collectively known as regulated products.
For alternative and herbal medicines specifically, NAFDAC implemented comprehensive regulations in 2019 covering registration, advertisement, and labeling. These regulations prohibit the manufacture, importation, exportation, distribution, advertisement, sale, or use of herbal medicines unless properly registered. The agency’s Registration, Advertisement, and Labelling Regulations mandate that products must pass safety and efficacy evaluations before market authorization.
However, the reality of enforcement tells a more complicated story. According to Ojonugwa and Gwom’s research, NAFDAC faces several interconnected challenges:
Corruption and Regulatory Capture: The research identified corruption as a significant hindrance to successfully achieving regulatory goals. Alternative medical practitioners frequently bribe government officials to evade arrest or compliance with regulatory standards—a practice that fundamentally undermines the entire regulatory framework.
Inadequate Penalties: Current penalties imposed on defaulters are insufficiently severe to deter potential offenders. This creates a cost-benefit calculation where the profits from selling substandard or counterfeit medicines outweigh the risks of detection and punishment.
Resource Constraints: The agency suffers from a lack of systems and manpower for proper monitoring. With Nigeria’s pharmaceutical market expanding rapidly—approximately 70% of medicines are imported—NAFDAC’s inspection capacity has struggled to keep pace with the volume of products entering the country.
Public Awareness Gap: A significant proportion of Nigeria’s population (37.98% according to 2021 literacy statistics) lacks the education necessary to identify counterfeit medicines or understand regulatory frameworks, making them particularly vulnerable to pharmaceutical fraud.
Technological Evolution: The never-ending evolution of internet commerce has reshaped counterfeit technology and trade networks. Research by Adigwe and colleagues found that 72.68% of pharmacists identified online drug commerce as contributing to counterfeit medicine circulation, yet regulatory frameworks have lagged behind these technological developments.
The human cost of these failures is staggering. Between 1995 and 1996, fake meningitis vaccines obtained from the Onitsha drug market killed thousands of children in Nigeria and neighboring Niger Republic. Research published in the African Journal of Traditional, Complementary, and Alternative Medicines indicates that prior to NAFDAC’s intensified efforts in the early 2000s, counterfeit drugs comprised up to 40% of the Nigerian pharmaceutical market.
The American Experience: AI-Powered Regulatory Innovation
While Nigeria struggles with fundamental enforcement challenges, the United States Food and Drug Administration (FDA) has embarked on an ambitious transformation of its regulatory apparatus through AI integration. This evolution offers valuable lessons for NAFDAC and Nigerian policymakers.
Strategic Framework and Action Plans
Research by Fu, Jia, Liu, Pang, and Cui published in Acta Pharmaceutica Sinica B in 2025 provides a comprehensive overview of AI applications in drug regulation globally, with particular attention to FDA initiatives. The FDA has developed a systematic approach to AI integration through multiple interconnected action plans:
Technology Modernization Action Plan (TMAP, 2019): This foundational strategy outlined the FDA’s approach to updating computer hardware, software, data, and analytics. The plan specifically aimed to establish a robust foundation for enhancing regulatory efficiency by integrating AI and other technologies to facilitate data usage in regulatory decision-making.
Data Modernization Action Plan (DMAP, 2021): Recognizing that the FDA’s data systems were primarily geared toward unstructured, non-digital, document-based information paradigms (such as PDFs), DMAP proposed a framework for transforming how the agency handles data. The plan emphasizes updated methods, information technology, and data usage processes to accelerate accessibility to better therapeutic products. Driver projects under DMAP explicitly aim to employ predictive models and cutting-edge technologies including AI.
Enterprise Modernization Action Plan (EMAP, 2022): This initiative focused on improving operational efficiency and enhancing data utilization by optimizing common and essential business processes—recognizing that technology alone cannot transform regulation without corresponding procedural innovation.
AI/ML SaMD Action Plan (2021-2024): Most significantly for pharmaceutical regulation, the FDA finalized its AI/ML-Based Software as a Medical Device Action Plan in December 2024. Research by Singh, Cheng, Kwan, and Ebinger published in Mayo Clinic Proceedings: Digital Health in 2025 explains that this framework introduces a Total Product Life Cycle (TPLC) regulatory approach specifically designed to address AI’s adaptive nature.
The AI/ML action plan incorporates “Good Machine Learning Practices”—ten principles guiding the development of high-quality and safe AI/ML devices. These principles encourage manufacturers to apply robust cybersecurity and coding practices, use unbiased and representative datasets, and maintain transparency with end-users about algorithm updates.
Practical Applications Delivering Results
The strategic vision has translated into concrete applications that demonstrate AI’s transformative potential:
Safety Surveillance and Pharmacovigilance: The FDA receives more than 2 million postmarket adverse event reports annually. Research by Fu and colleagues documents how the Center for Drug Evaluation and Research (CDER) has developed machine learning models to automate the classification of adverse event reports in the FDA Adverse Event Reporting System (FAERS). The Office of Surveillance and Epidemiology developed classification models using supervised machine learning and text engineering methods, streamlining the review process and allowing expert evaluation to focus on more informative cases.
The FDA’s Sentinel System, fully operationalized in 2016, represents a landmark achievement in proactive safety surveillance. According to the Sentinel System Five-Year Strategy 2019-2023, the system now leverages natural language processing (NLP) and machine learning to automatically extract features from unstructured electronic health records to support postmarket safety analysis. The Biologics Effectiveness and Safety System (BEST), developed in 2017, similarly employs NLP technology to automatically report adverse events for biologics.
Workflow Optimization: The FDA’s Office of Pharmaceutical Quality developed the Knowledge-aided Assessment and Structured Application (KASA) tool in 2018 to address the increasing volume of Abbreviated New Drug Applications. Unlike text-based electronic Common Technical Documents (eCTD), KASA uses structured assessment templates and risk-level-based algorithms, significantly improving review efficiency, objectivity, and consistency. By 2020, KASA had expanded from solid oral dosage forms to liquid dosage forms and has since been extended to new drug and biologics applications.
The Computerized Labeling Assessment Tool (CLAT), developed in 2020, employs AI technology to automatically review labels and identify elements including prescription information, package boxes, and container labels. Additionally, the FDA developed an AI-based component in 2020 allowing querying of labeling documents using customized and fine-tuned language models, enhancing the specificity and relevance of retrieved labeling data for drug repurposing studies.
Regulatory Science Research: The FDA’s National Center for Toxicological Research launched the AI4TOX program, applying advanced AI methods to develop tools supporting regulatory science and strengthening safety reviews of FDA-regulated products. This program includes initiatives like AnimalGAN for predicting animal toxicology data, SafetAI for developing deep learning methods for toxicological endpoints, and DeepDILI for predicting drug-induced liver injury.
Measured Impact and Ongoing Challenges
Research by Muehlematter, Bluethgen, and Vokinger published in The Lancet Digital Health in 2023 analyzed FDA-cleared AI/ML-based medical devices between 2019 and 2021, providing critical insights into both successes and ongoing challenges. Their study of 285 AI/ML-based medical devices found that 237 (83.2%) were cleared for radiology, 26 (9.1%) for cardiovascular use, and smaller numbers for other specialties.
Importantly, the research revealed potential concerns about the 510(k) predicate pathway used for many AI/ML device clearances. More than a third (32.6%) of cleared AI/ML-based medical devices originated from non-AI/ML-based devices in the first generation, and devices could be traced back over a mean of 1.5 generations until a de novo or non-AI/ML-based predicate device was reached. For radiology devices specifically, AI/ML tasks changed in approximately every second device connection (47.8%), raising questions about the appropriateness of substantial equivalence determinations when underlying technologies differ significantly.
Despite these concerns, the FDA’s systematic approach has yielded tangible improvements in regulatory efficiency and scientific rigor. Research by Warraich, Tazbaz, and Califf published in JAMA in 2025 notes that over 100 drug development applications utilizing AI/ML technologies have been submitted to the FDA, demonstrating industry confidence in the regulatory framework for AI-enabled innovations.
Translating Lessons for Nigeria: A Roadmap for NAFDAC
The American experience offers crucial insights for Nigerian policymakers seeking to modernize NAFDAC’s regulatory capabilities. However, direct transplantation of FDA approaches would ignore Nigeria’s distinct challenges and resource constraints. Instead, policymakers should consider a phased, contextually appropriate adoption strategy.
Phase 1: Foundation Building (Years 1-2)
Data Infrastructure Development: Before implementing sophisticated AI solutions, NAFDAC requires fundamental data modernization. Current regulatory submissions to NAFDAC, like those to the FDA historically, rely heavily on unstructured documents. Investment in converting to structured, machine-readable formats must be prioritized.
The Chinese experience offers relevant precedents. Research by Fu and colleagues notes that China’s National Medical Products Administration (NMPA) issued the “Action Plan for Accelerating Smart Drug Regulation” in 2019, emphasizing cooperation with enterprises and third-party organizations to build regulatory data analysis laboratories. Nigeria could similarly leverage partnerships with pharmaceutical manufacturers, universities, and technology companies to establish data infrastructure without bearing the entire financial burden alone.
Pilot Projects in High-Impact Areas: Rather than attempting wholesale transformation, NAFDAC should identify specific high-value applications for initial AI deployment. Based on the challenges identified by Adigwe and colleagues, three areas warrant prioritization:
- Counterfeit Detection Systems: Given that 72.68% of pharmacists identified online drug commerce as contributing to counterfeit circulation, AI-powered systems for monitoring online pharmaceutical sales represent a high-impact intervention. Natural language processing could scan online marketplaces, social media platforms, and websites for illicit pharmaceutical sales, automatically flagging suspicious activity for human review.
- Inspection Optimization: With 90.93% of pharmacists citing inadequate inspection as a challenge, AI could help NAFDAC allocate its limited inspection resources more effectively. Machine learning models could analyze historical inspection data, import records, adverse event reports, and other data sources to generate risk scores for facilities and importers, prioritizing inspections where risks are highest.
- Adverse Event Signal Detection: Following the FDA’s FAERS model, NAFDAC could implement machine learning systems to automatically classify and triage adverse event reports, identifying potential safety signals faster than manual review allows.
Capacity Building: The most sophisticated technology provides no value without personnel capable of using it effectively. Research by Ojonugwa and Gwom emphasizes the need for highly trained interdisciplinary talent bridging data science, computer science, and pharmaceuticals. NAFDAC should establish partnerships with Nigerian universities to develop specialized training programs, potentially including scholarships for pharmacy and computer science students interested in regulatory science careers.
Phase 2: Expansion and Integration (Years 3-5)
Regulatory Framework Modernization: Drawing from the FDA’s TPLC approach for AI/ML devices, NAFDAC should develop specific guidance for AI-enabled pharmaceutical products and regulatory technologies. This framework must address:
- Standards for AI/ML validation in pharmaceutical applications
- Requirements for transparency and explainability of AI systems used in drug development or safety monitoring
- Procedures for updating and maintaining AI systems while ensuring continued regulatory compliance
- Ethical guidelines for AI deployment, particularly regarding data privacy and algorithmic bias
Cross-Border Collaboration Systems: Given that 90.43% of pharmacists identified poor cross-border enforcement as a challenge and approximately 70% of Nigerian medicines are imported, AI could facilitate regional regulatory cooperation. The European Medicines Agency’s (EMA) approach offers relevant lessons. Research by Fu and colleagues documents how the EMA established the Big Data Steering Group and Analytics Centre of Excellence to integrate big data and AI into regulatory evaluation and decision-making across the European Medicines Regulatory Network.
Nigeria could lead development of an ECOWAS (Economic Community of West African States) pharmaceutical regulatory data-sharing platform, using AI to identify suspicious patterns in cross-border pharmaceutical trade. Machine learning models could detect anomalies such as unusually large shipments, products transiting through multiple countries before reaching Nigeria, or importers with histories of regulatory violations.
Intelligent Document Processing: Building on the FDA’s experience with KASA and CLAT, NAFDAC could implement AI-powered systems for processing registration applications. Natural language processing could extract key information from application documents, automatically populate regulatory templates, identify missing information, and flag potential issues for human review—dramatically accelerating the registration process while maintaining or improving review quality.
Phase 3: Advanced Capabilities (Years 5+)
Predictive Analytics for Public Health: Advanced AI systems could integrate diverse data sources—import records, manufacturing inspections, adverse event reports, disease surveillance data, social media monitoring, and more—to generate predictive insights about emerging pharmaceutical safety threats. For instance, machine learning models might identify geographic clusters of adverse events suggesting counterfeit medicine circulation before traditional surveillance systems would detect the pattern.
Automated Surveillance of Alternative Medicine: Research by Ojonugwa and Gwom documents NAFDAC’s extensive regulatory framework for alternative and herbal medicines, including detailed requirements for registration, labeling, and advertising. However, enforcement remains challenging given the informal nature of much traditional medicine practice. Computer vision systems could monitor marketplace images to identify unregistered herbal products, while NLP could scan social media and websites for prohibited health claims.
Real-World Evidence Integration: Following the FDA’s Sentinel Initiative model, NAFDAC could develop systems integrating electronic health records, pharmacy dispensing data, and other real-world data sources to conduct post-market safety surveillance. Machine learning algorithms could identify safety signals from routine clinical practice data, complementing spontaneous adverse event reporting systems.
Addressing the Elephant in the Room: Practical Constraints
The vision outlined above requires candid acknowledgment of significant obstacles. Resource constraints represent the most obvious challenge. The FDA’s technology modernization efforts have benefited from substantial Congressional appropriations through the Prescription Drug User Fee Act (PDUFA), Biosimilar User Fee Act (BsUFA), and Generic Drug User Fee Amendments (GDUFA). NAFDAC operates under far more constrained budgets.
However, several factors suggest AI adoption may be more financially feasible than initially apparent:
Decreasing Technology Costs: Cloud computing services and open-source AI tools have dramatically reduced the capital requirements for deploying AI systems. NAFDAC need not build massive data centers or license expensive proprietary software—cloud-based solutions allow pay-as-you-go models that scale with organizational capacity.
Multiplier Effects: Even modest AI deployments can generate efficiency gains that free resources for expanded implementation. If AI-powered document processing reduces registration review times by 30%, the staff hours saved can be redirected to implementing additional AI applications or strengthening other regulatory functions.
International Support: Development partners including the World Health Organization, World Bank, and bilateral development agencies have expressed interest in supporting pharmaceutical regulatory strengthening in developing nations. AI-focused initiatives could attract funding more readily than traditional capacity-building programs, given international interest in technological innovation.
Private Sector Partnership Models: The NMPA’s approach of cooperating with enterprises and third-party organizations offers relevant precedents. Pharmaceutical manufacturers and technology companies have vested interests in efficient, science-based regulation. Public-private partnerships could share AI development costs while maintaining NAFDAC’s independence in regulatory decision-making.
The corruption challenge documented by Ojonugwa and Gwom requires equally direct confrontation. AI systems, properly designed, could actually reduce corruption opportunities by replacing discretionary human judgments with algorithm-driven decisions in appropriate contexts. For instance, risk-based inspection scheduling algorithms could eliminate situations where inspectors select targets based on personal relationships rather than objective risk criteria.
However, AI can also create new corruption vulnerabilities if systems lack transparency or if gatekeepers controlling AI systems demand bribes for favorable algorithmic outcomes. Mitigating these risks requires:
- Algorithmic transparency, with clear documentation of how AI systems make decisions
- Regular audits of AI system outputs to detect anomalous patterns suggesting manipulation
- Multiple validation layers combining AI recommendations with human oversight
- Robust data governance preventing unauthorized access or modification of system inputs
- Whistleblower protections for staff reporting corruption involving AI systems
Policy Recommendations
Based on this analysis, Nigerian policymakers should consider the following concrete actions:
- Establish a NAFDAC Technology Modernization Task Force charged with developing a comprehensive 5-year plan for AI integration, modeled on the FDA’s TMAP but adapted to Nigerian contexts. This task force should include representation from NAFDAC, pharmaceutical manufacturers, healthcare providers, academic institutions, technology companies, and civil society organizations.
- Appropriate dedicated funding for regulatory technology modernization, potentially through earmarked fees on pharmaceutical registrations and imports. A modest fee increase generating $5-10 million annually could support substantial AI capacity development while remaining financially sustainable.
- Develop ECOWAS-level coordination mechanisms for pharmaceutical regulation, with Nigeria taking a leadership role given NAFDAC’s regional reputation. Harmonized regulatory standards and data-sharing agreements would multiply the benefits of AI investments while addressing cross-border enforcement challenges.
- Partner with Nigerian universities to establish specialized graduate programs in pharmaceutical regulatory science with AI/data science components. These programs could create the talent pipeline NAFDAC requires while strengthening Nigeria’s broader technology ecosystem.
- Mandate structured, machine-readable formats for regulatory submissions on a phased timeline, beginning with new drug registrations and gradually expanding to all submission types. This data infrastructure is essential for any AI application.
- Establish clear ethical guidelines and transparency requirements for AI use in pharmaceutical regulation, ensuring algorithmic accountability and preventing bias against local manufacturers or particular therapeutic categories.
- Create regulatory sandboxes allowing controlled experimentation with AI technologies in pharmaceutical development and regulation, similar to the FDA’s ISTAND pilot program. This would encourage innovation while maintaining appropriate oversight.
Conclusion: The Imperative of Modernization
The pharmaceutical regulatory challenges facing Nigeria are not unique, but their intensity and human consequences demand urgent attention. Research by Adigwe and colleagues found that 97.2% of Nigerian pharmacists agree counterfeit medicines pose serious threats to healthcare, and 99.5% associate counterfeit pharmaceuticals with treatment failure. The sub-Saharan African region records approximately 280,000 child deaths annually from counterfeit antimalarial and pneumonia medications. These are not abstract regulatory failures—they represent children who will never reach adulthood, parents buried prematurely, and communities devastated by preventable tragedies.
The FDA’s experience demonstrates that AI can meaningfully improve pharmaceutical regulation—accelerating review processes, enhancing safety surveillance, optimizing resource allocation, and ultimately protecting public health more effectively. These are not theoretical benefits; they reflect documented outcomes from systems already operational.
For Nigerian policymakers, the question is not whether NAFDAC should embrace AI, but rather how quickly and effectively this transformation can occur. The agency’s challenges—corruption, resource constraints, inadequate legislation, poor inter-agency collaboration—are formidable but not insurmountable. Indeed, properly designed AI systems can help address several of these obstacles while improving regulatory efficiency and effectiveness.
The path forward requires political will, strategic investment, international partnership, and sustained commitment to evidence-based policymaking. It requires acknowledging that 20th-century regulatory approaches cannot adequately protect public health in 21st-century pharmaceutical markets characterized by global supply chains, online commerce, and rapid technological evolution.
Most fundamentally, it requires recognizing that pharmaceutical regulation is not a cost to be minimized but an investment in human capital, economic development, and national security. Every child saved from counterfeit antimalarials, every family protected from substandard medicines, represents both a moral imperative and an economic benefit as Nigeria seeks to build a healthier, more productive society.
The tools exist. The roadmap is clear. The only question is whether Nigerian policymakers will seize this opportunity to transform NAFDAC into a modern, AI-enabled regulatory agency capable of protecting public health in an increasingly complex pharmaceutical landscape. The stakes—measured in human lives—could not be higher.
