Artificial intelligence is emerging as a critical solution to the longstanding labor shortage in rare disease research and treatment. Despite advances in gene editing and drug design, thousands of rare diseases remain without effective therapies—largely due to a lack of skilled scientists and the high cost and time required to develop treatments. Companies like Insilico Medicine and GenEditBio are turning to AI to overcome these barriers, using intelligent systems to accelerate discovery and scale innovation. At Web Summit Qatar, Insilico Medicine’s CEO and founder Alex Aliper outlined the company’s vision of “pharmaceutical superintelligence.” The goal is to create a multi-modal, multi-task AI capable of handling diverse drug discovery challenges with superhuman precision. To achieve this, Insilico launched its MMAI Gym—a platform designed to train general-purpose large language models like ChatGPT and Gemini to match or exceed the performance of specialized models in biotech tasks. These AI systems can analyze biological, chemical, and clinical data to generate hypotheses about disease targets, design novel molecules, and even repurpose existing drugs. For example, Insilico recently used its AI to identify potential treatments for ALS, a rare and devastating neurological disorder. By automating labor-intensive processes traditionally requiring large teams of chemists and biologists, the company has drastically reduced the time and cost of early-stage drug discovery. Aliper emphasized that AI is essential to increasing productivity in pharmaceutical research, especially given the vast number of untreatable rare diseases and the global shortage of talent. Beyond drug discovery, the challenge of delivering therapies to the right cells remains a major hurdle. GenEditBio is tackling this with a next-generation approach to CRISPR gene editing that focuses on in vivo delivery—directly injecting gene-editing tools into the body. The company’s proprietary ePDV (engineered protein delivery vehicle) is a virus-like particle designed to safely transport gene-editing machinery to specific tissues such as the eye, liver, or nervous system. Using AI, GenEditBio analyzes vast libraries of nonviral, nonlipid polymer nanoparticles to identify which structures can efficiently and safely deliver payloads. The AI predicts how chemical changes affect targeting and immune response, then tests these predictions in wet labs. Results are fed back into the system to refine future predictions, creating a powerful feedback loop. This approach, the company says, reduces costs, improves scalability, and enables off-the-shelf treatments that could be accessible globally. Despite progress, both companies acknowledge a persistent bottleneck: data. High-quality, diverse biological data is essential for training accurate AI models, yet most datasets are skewed toward populations in the Western world. Aliper stressed the need for more inclusive, locally generated data to ensure AI systems can serve all patients equitably. Insilico’s automated labs are helping by generating rich, multi-layered biological data from disease samples at scale, while GenEditBio leverages natural evolutionary patterns encoded in non-coding DNA—information long ignored but now accessible to AI. The ultimate goal, according to Aliper, is the development of digital twins—virtual human models that can simulate disease progression and test treatments in silico. These virtual clinical trials could dramatically speed up drug development and personalize care. With only around 50 new drugs approved by the FDA annually, Aliper believes AI is key to breaking through the current plateau. As global populations age and chronic diseases rise, he hopes that in the next 10 to 20 years, AI-driven innovation will unlock a new era of therapeutic options for patients with rare and complex conditions.