During a flight from London to Nairobi, I fell into a deep, sedation-like sleep due to a medication I take for glossopharyngeal neuralgia. Thanks to a misunderstanding with airport authorities over this prescribed medication, staff mistook me for a drug mule. I awoke to sniffer dogs and police surrounding me, and it was only after they called my doctor to confirm my medication and condition—and after the dogs found no illicit drugs on me or in my luggage—that the situation was finally resolved.

This is just one of the many trials I face from living with a disease so rare, it affects about 2 people per million.

Glossopharyngeal neuralgia is a condition characterised by severe, stabbing pain in the ear, base of the tongue, tonsil, or beneath the angle of the jaw. Physically, I appear ‘normal,’ making it difficult for others to understand the excruciating pain I experience every few minutes. This disconnect between outward appearance and internal suffering is a common struggle for many individuals with rare diseases.

Conventional treatments have ranged from anticonvulsants to invasive surgeries such as microvascular decompression or Gamma Knife radio-surgery. While these treatments can be effective, they come with significant risks and side effects, and for most, they eventually lose their efficacy. What if instead, a tiny AI-driven microchip could regulate nerve activity with precision and safety and eliminate the pain?

AI has the potential to transform treatment of rare diseases like mine, if given the chance.

Approximately 7,000 rare diseases collectively impact tens of millions of people. Yet less than five per cent of these conditions, which range from rare cancers to genetic disorders affecting children, have a U.S. Food and Drug Administration-approved treatment. The problem is not that these diseases are inherently untreatable; rather, it’s that research and development efforts have been insufficient.

New tools exist that could revolutionise this landscape. Artificial intelligence and advanced computational methods have the potential to accelerate diagnosis, drug discovery and development for rare diseases. Take the case of glossopharyngeal neuralgia. An AI-based system could regulate the electrical stimulation at the root of the pain, mirroring the effects of anti epileptic drugs but without the side effects or risks of surgery.

For a disease like cystic fibrosis, as another example, AI could identify new genetic mutations and predict patient-specific responses to treatments, enabling personalised therapies that significantly improve outcomes. And for the rare disease Duchenne muscular dystrophy, it could analyse vast amounts of data to discover novel drug candidates, accelerating the development of treatments that could slow down or even halt the progression of the disease.

Most importantly, once developed, such AI systems could be deployed globally by governments and healthcare providers, supported by international organisations like the World Health Organization (WHO) and the United Nations, ensuring that even patients in remote or underserved areas can benefit from these cutting-edge advancements.

This global scalability has the potential to democratise access to high-quality care for rare disease patients worldwide, bridging the gap between specialised medical knowledge and those who need it most, regardless of their geographical location or economic status.​​

Yet, despite these advantages, the integration of AI in treating rare diseases has been slow, as has investment in rare diseases in general. The primary challenges lie in the economics of drug development. Pharmaceutical companies often hesitate to invest in treatments for rare diseases due to the limited potential patient base and uncertain return in investment. Indeed, diseases affecting a small fraction of the population attract less attention and funding from the research community and commercial investors.

Their complexity requires significant upfront investment in research and trials, and enormous quantities of uniformly collected and shareable high-quality data. Additionally, research efforts are often siloed, with proprietary information making it difficult to share crucial data needed to identify solutions.

The question then arises: how can we overcome these challenges? The answer partly lies in visibility and advocacy. Rare diseases are often overshadowed by more common conditions that promise a greater return on investment for pharmaceutical companies and researchers. So, those of us living with diseases and our families must make these diseases more visible. Drug development enterprises would also need to change, to be less of a competition and more of collaboration.

While it may be hard to address the variations of 7,000 rare diseases, it’s important to try. By focusing on rare diseases, AI developers can not only showcase the versatility and transformative power of their technologies but also underscore their commitment to enhancing human life, regardless of the economic or publicity gains. This is the future of healthcare—a future where no condition is too rare to merit attention and no patient too isolated to receive care.

As we advance in technology, let us not forget those who are silently suffering from rare diseases. There may be fewer of us per disease, but we deserve attention and innovation, too. AI has the potential to bring significant relief to those who have few options left, and it is our moral imperative to ensure these advancements reach all corners of the healthcare universe, no matter how rare the condition.

Maxwell Gomera is the Resident Representative of the United Nations Development Program in South Africa and Director of the Africa Sustainable Finance Hub.