Yu-Xiu Lin, M.Sc & Thai-Yen Ling, Ph.D.
Graduate Institute of Pharmacology, College of Medicine, National Taiwan University, Taiwan

Artificial intelligence (AI) is reshaping the life sciences industry at an unprecedented pace. Among the most transformative areas is cell and gene therapy (CGT), a field that promises curative approaches for cancer, genetic disorders, autoimmune diseases, and degenerative conditions. As CGT advances from experimental innovation to commercial reality, AI is emerging not merely as a supporting technology but as a central driver of scalability, efficiency, and precision. Recent market analyses suggest that AI-powered digital innovation is poised to redefine the manufacturing and development landscape of advanced therapies over the coming decade. The global cell and gene therapy manufacturing market, valued at approximately USD 14.69 billion in 2025, is projected to surpass USD 122.86 billion by 2034, reflecting a compound annual growth rate of 26.6% (BioSpace, 2024). Such rapid expansion is unlikely to be sustainable without digital optimization, making AI-driven systems increasingly indispensable in addressing biological variability, manufacturing bottlenecks, and regulatory complexity inherent to advanced therapies.

One of the most significant applications of AI lies in intelligent manufacturing. Cell therapy production involves dynamic biological processes that are highly sensitive to environmental fluctuations, culture conditions, and operator variability. Machine learning algorithms now enable continuous monitoring of cell growth kinetics, metabolic indicators, and environmental parameters in real time. By identifying deviations at early stages, AI reduces batch failures and improves consistency across production runs. This transition from reactive quality control to predictive process management represents a fundamental shift in biomanufacturing philosophy (BioSpace, 2024). In addition, AI supports predictive quality modeling. Whereas traditional quality assurance often relies on endpoint testing that detects failure only after significant resources have been expended, AI systems trained on historical production datasets can recognize patterns associated with suboptimal outcomes and intervene proactively. This approach enhances reproducibility and lowers overall production costs, an essential advancement for therapies that are frequently patient-specific and manufactured in small batches. Digital twin technology further extends this paradigm. By constructing virtual replicas of manufacturing environments, AI-driven simulations allow researchers to test parameter adjustments, forecast yields, and assess risk scenarios before implementing changes in physical systems, thereby accelerating process optimization and shortening the path from laboratory development to commercial-scale production (BioSpace, 2024).

Beyond manufacturing, AI is equally transformative in early-stage research and therapeutic design. In CAR-T and other engineered cell therapies, machine learning models analyze genomic, transcriptomic, and proteomic datasets to identify optimal targets and refine receptor specificity. Predictive modeling assists in minimizing off-target toxicity and improving therapeutic durability (McKinsey & Company, 2023). Deep learning approaches are also being explored to anticipate cytokine release syndrome risk and other adverse immune reactions, potentially increasing clinical trial success rates (Ni, 2025). The integration of AI with computational immunology and protein engineering represents another frontier. Generative models trained on protein sequence data are now capable of assisting in the design of T-cell receptors and antibody constructs with enhanced binding properties (Xie, 2026). These technologies expand the conceptual boundaries of cell therapy development, shifting the field from empirical experimentation toward data-guided design.

AI is also enabling more personalized therapeutic strategies. By integrating multi-omic patient data with clinical histories and treatment outcomes, predictive models can support individualized therapy selection and optimization (Choudhery, 2024). Such tools are particularly relevant in autologous cell therapies, where patient heterogeneity directly influences product characteristics and therapeutic response. Increasing emphasis is also being placed on the integration of real-world data (RWD) into AI-driven cell therapy platforms. Post-approval treatment outcomes, long-term safety monitoring, and real-world effectiveness metrics provide insights that extend beyond traditional clinical trial endpoints. When systematically analyzed, these datasets can inform continuous refinement of manufacturing parameters, product characterization strategies, and therapeutic design. AI systems capable of processing large-scale real-world evidence (RWE) offer the potential for adaptive optimization across the entire therapy lifecycle. By linking longitudinal clinical performance back to specific production variables, AI enables a feedback loop between bedside outcomes and biomanufacturing processes. This bidirectional data flow may ultimately shift cell therapy from a static, one-time product paradigm toward a continuously learning and evolving therapeutic ecosystem.

The geographic and industrial landscape reflects these technological shifts. North America currently dominates the CGT manufacturing market, supported by strong infrastructure and regulatory frameworks, while the Asia-Pacific region is projected to experience the fastest growth, driven by expanding investments in digital biomanufacturing platforms and advanced therapy facilities (BioSpace, 2024). Contract manufacturing organizations are increasingly incorporating AI into their production systems to enhance competitiveness and scalability. Despite its transformative potential, AI integration faces important challenges. Data standardization remains a critical barrier, as manufacturing datasets are often fragmented or inconsistently formatted, restricting model generalizability (Di Cerbo, 2025). Regulatory oversight also presents complexity, requiring AI-driven decision systems to demonstrate transparency, reproducibility, and explainability. Furthermore, the capital investment required for automation infrastructure and computational platforms may limit adoption among smaller biotechnology firms.

Nevertheless, the trajectory is clear. As computational models become more sophisticated and data ecosystems more integrated, AI will continue to reshape how advanced therapies are designed, manufactured, and delivered. Multi-modal AI systems that combine imaging analytics, omics integration, real-world evidence, and process engineering data are likely to define the next phase of digital biomanufacturing. Artificial intelligence is no longer an auxiliary innovation in cell therapy; it is becoming the digital backbone of next-generation advanced therapeutics. By enabling predictive manufacturing, intelligent therapy design, lifecycle learning, and personalized treatment strategies, AI is accelerating the transition of cell therapy from specialized innovation to scalable and adaptive clinical reality.

References

1. BioSpace. (2024). AI-powered cell and gene therapy manufacturing market outlook 2034: Scaling advanced therapies with digital innovation. https://www.biospace.com/press-releases/ai-powered-cell-and-gene-therapy-manufacturing-market-outlook-2034-scaling-advanced-therapies-with-digital-innovation
2. McKinsey & Company. (2023). How AI can accelerate R&D for cell and gene therapies. https://www.mckinsey.com/industries/life-sciences/our-insights/how-ai-can-accelerate-r-and-d-for-cell-and-gene-therapies.
3. Ni, Y., Zhu, L., & Li, S. (2025). Bio AI Agent: A Multi-Agent Artificial Intelligence System for Autonomous CAR-T Cell Therapy Development with Integrated Target Discovery, Toxicity Prediction, and Rational Molecular Design. arXiv preprint arXiv:2511.08649.
4. Xie, L., Pelissier, A., Shao, Y., & Martinez, M. I. R. (2026). AI Developments for T and B Cell Receptor Modeling and Therapeutic Design. arXiv preprint arXiv:2601.17138.
5. Choudhery, M. S., Arif, T., & Mahmood, R. (2025). Applications of artificial intelligence in stem cell therapy. World Journal of Stem Cells, 17(8), 106086.
6. Di Cerbo, V., Song, H. W., Herbst, L., Hart, S. J., Ladi, R., Jiang, S., … & Sethi, D. (2025). Artificial intelligence, machine learning, and digitalization systems in the cell and gene therapy sector: a guidance document from the ISCT industry committees. Cytotherapy, 27(8), 903-909.