Modern biotechnology possesses the tools to edit genes and design drugs, yet thousands of rare diseases remain without treatment. According to executives from Insilico Medicine and GenEditBio, the persistent missing ingredient has been a shortage of skilled people to advance the work. They state that artificial intelligence is now becoming the essential force multiplier, enabling scientists to tackle problems the industry has long overlooked.
Speaking at Web Summit Qatar, Insilico Medicine’s CEO and founder Alex Aliper detailed his company’s goal to develop what he calls “pharmaceutical superintelligence.” The company recently launched its “MMAI Gym,” which aims to train generalist large language models to perform as well as specialized models. The objective is to build a multi-modal, multi-task model that can solve many different drug discovery tasks simultaneously with superhuman accuracy.
Aliper emphasized the need for this technology to increase the pharmaceutical industry’s productivity and address the critical shortage of labor and talent. He noted that thousands of diseases, including many rare disorders, still lack any treatment options. Insilico’s platform integrates biological, chemical, and clinical data to generate hypotheses about disease targets and candidate molecules. By automating steps that once required large teams of scientists, the company says it can sift through vast design spaces, nominate high-quality therapeutic candidates, and repurpose existing drugs at dramatically reduced cost and time. For instance, Insilico recently used its AI models to identify existing drugs with potential for repurposing to treat ALS, a rare neurological disorder.
However, the labor bottleneck extends beyond drug discovery. Even when AI identifies promising targets or therapies, many diseases require interventions at a more fundamental biological level. GenEditBio is part of a “second wave” of CRISPR gene editing, moving the process from editing cells outside the body toward precise delivery inside the body. The company’s goal is to make gene editing a one-time injection directly into affected tissue.
GenEditBio’s co-founder and CEO Tian Zhu explained that the company has developed a proprietary engineered protein delivery vehicle, which is a virus-like particle. They use AI and machine learning to mine natural resources, studying viruses that have an affinity for specific tissue types. This refers to the company’s massive library of unique, nonviral polymer nanoparticles designed to safely transport gene-editing tools into specific cells.
The company’s NanoGalaxy platform uses AI to analyze data and identify how chemical structures correlate with specific tissue targets, such as the eye, liver, or nervous system. The AI then predicts which modifications to a delivery vehicle’s chemistry will help it carry its payload without triggering an immune response. GenEditBio tests these delivery vehicles in live labs, and the results are fed back into the AI to refine its predictive accuracy. Zhu argues that this efficient, tissue-specific delivery is a prerequisite for in vivo gene editing, reducing costs and standardizing a historically difficult process to scale. She compares it to an off-the-shelf drug that works for multiple patients, increasing affordability and global access. The company recently received FDA approval to begin trials of a CRISPR therapy for corneal dystrophy.
Progress in AI-driven biotech ultimately encounters a persistent data problem. Modeling the complexities of human biology requires far more high-quality data than researchers currently have. Aliper pointed out that the corpus of medical data is heavily biased toward the Western world where it is generated. He advocates for more local efforts to create a more balanced set of original data so that AI models can be more capable. Insilico’s automated labs generate multi-layer biological data from disease samples at scale, feeding it directly into its discovery platform.
Zhu believes the necessary data already exists within the human body, shaped by millennia of evolution. She notes that only a small fraction of DNA codes for proteins, while the rest acts as an instruction manual for gene behavior—information historically difficult for humans to interpret but increasingly accessible to AI. In the lab, GenEditBio tests thousands of delivery nanoparticles in parallel, creating data sets it calls “gold for AI systems” to train its models and support collaborations.
Looking ahead, Aliper sees one of the next major efforts as building digital twins of humans to run virtual clinical trials, a field he describes as still in its infancy. He observed that the number of new drugs approved by the FDA annually has plateaued at around fifty, while the rise in chronic disorders grows as the global population ages. His hope is that in ten to twenty years, there will be many more therapeutic options for the personalized treatment of patients.