The increasing uptake of artificial intelligence (AI) by hospitals and telemedicine providers is introducing a new layer of risk for insurers, including those operating in Africa, raising difficult questions over liability when AI-assisted diagnoses result in medical errors.
Recent reports by global health and insurance organisations, which African insurers may want to follow with interest, highlight widening liability gaps and the need for clear frameworks on accountability, risk transfer and patient protection in AI-assisted care.
While AI promises to expand access and improve diagnostic speed across underserved markets, it is simultaneously blurring the lines of medical responsibility between clinicians, technology providers and healthcare institutions.
For insurers and risk managers, this shift presents a growing underwriting challenge. As healthcare delivery becomes more digital, traditional policy structures are being stretched beyond their original design, forcing the industry to rethink how risk is priced, shared and ultimately absorbed in an increasingly AI-enabled care ecosystem.
A report by the Artificial Intelligence Underwriting Company (AIUC), co-authored with researchers from Anthropic and OpenAI, said more than 90% of insurers’ exposure to AI-related risks is embedded in conventional insurance policies that were not designed for increasingly autonomous AI systems.
The report, titled ‘Underwriting the Agent Economy’, warned that AI-related litigation is expanding beyond chatbots to autonomous systems capable of making decisions, exposing businesses to claims, “ranging from professional negligence to cyber attacks and wrongful death”.
“Businesses cannot adopt AI unless they know the risk has been quantified and managed,” Rajiv Dattani, co-founder of AIUC, said, adding that insurers must develop products specifically designed to cover AI-related risks rather than relying on conventional policies.
The development comes as the Organisation for Economic Co-operation and Development (OECD) warned that while AI could transform healthcare delivery, regulatory and governance gaps continue to slow its safe deployment.
In its 2026 report, Scaling Artificial Intelligence in Health, the OECD said, “Artificial intelligence holds significant potential for the healthcare system. That potential is not being fully realised due to fragmented data foundations, non-aligned policies and practices, and structural and governance barriers to scalability.”
The organisation added that AI adoption remains constrained by “regulatory uncertainty” and gaps in governance despite its growing use across healthcare systems.
Similarly, the World Health Organisation (WHO) said hospitals are deploying AI to assist with disease diagnosis, medical imaging, patient triage, and telemedicine faster than governments are developing legal safeguards to determine responsibility when technology fails.
In its 2025 report, Artificial Intelligence in Health: State of Readiness across the WHO European Region, the organisation said, “Less than one in 10 countries have liability standards for AI in health, which determine who is responsible if an AI system makes an error or causes harm.”
WHO regional advisor on data, AI and digital health, Dr David Novillo Ortiz, said, “Without clear legal standards, clinicians may be reluctant to rely on AI tools and patients may have no clear path for recourse if something goes wrong.
“That’s why WHO/Europe urges countries to clarify accountability, establish redress mechanisms for harm, and ensure that AI systems are tested for safety, fairness, and real-world effectiveness before they reach patients,” he said.
The OECD also noted that while AI is widely used for administrative functions across member countries, only a small proportion of medical AI applications have been deployed at the national scale due to concerns regarding governance, transparency, patient safety, and accountability.
The reports come amid growing adoption of AI-enabled telemedicine platforms and diagnostic tools by healthcare providers worldwide, including Africa, prompting insurers to reassess whether existing health insurance, medical malpractice, and professional indemnity policies adequately cover claims arising from AI-assisted clinical decisions.
Industry analysts have noted that the absence of clear liability rules could lead to disputes among hospitals, physicians, software developers, and insurers whenever patients suffer injuries linked to AI-supported diagnosis or treatment, underscoring the need for dedicated AI liability frameworks as healthcare systems become increasingly digital.
Echoing these concerns in the Nigerian context, postdoctoral researcher at the Department of Pediatrics, Cumming School of Medicine, University of Calgary, Canada, Dr Babatope Oluwadamilare Adebiyi, argued in his paper, Governing Clinical AI in Nigeria: Accountability Gaps, Regulatory Pathways, and the Case for Named Decision Ownership, that the country’s regulatory framework has yet to address accountability for AI-driven clinical decisions.
It stated, “Nigeria’s clinical AI governance gap is real, significant, and growing. No existing legislation, regulation, or professional guidance clearly establishes who is accountable when a clinical AI system in a Nigerian hospital causes patient harm.
“The Federal Ministry of Health should develop a National Clinical AI Governance Policy establishing minimum standards for AI deployment in health facilities, including named decision ownership, validation standards, and patient rights provisions.”
Imole Latona, a Nigeria-based healthcare practitioner with three years of clinical experience, said healthcare professionals are already accustomed to managing medical risks through professional indemnity insurance, but the growing use of AI presents an entirely new layer of legal and ethical uncertainty.
“Healthcare systems recognise that adverse outcomes may occur even when clinicians act competently and compassionately. For that reason, hospitals and practitioners often maintain professional indemnity or medical malpractice insurance. Such protection exists not to excuse negligence but to provide financial cover where claims arise despite adherence to accepted standards of care, while ensuring patients who genuinely suffer avoidable harm have a path to compensation,” he said.
“This is where AI raises difficult questions. If an AI system contributes to a harmful clinical decision, who ultimately bears responsibility? The clinician, the hospital, the software developer, the manufacturer, or the regulator? AI holds extraordinary promise as a tool for diagnosis, research, education, and decision support, but it carries no professional licence, answers to no disciplinary council, and cannot stand before a patient or a court.
“That is why the future of healthcare cannot be built on AI alone. It must remain anchored in human professionals who are ethically accountable, emotionally present, legally responsible, and guided by the standards of their disciplines. AI should strengthen clinical judgment, bridge knowledge gaps, and expand access to care, but it should never replace the conscience, compassion, and responsibility that define the practice of medicine. Only then can innovation truly serve humanity rather than merely impress it.”
A United Nations Development Programme (UNDP) official said in a 2022 report on lessons from Covid-19 that insurance could be a solution for future crises. The piece argued that a more enabling regulatory environment is essential to support innovative health insurance models, with safeguards for data privacy, patients’ rights, and risk-based regulation for telemedicine.
COMMENTS