Foundational AI In Cell and Gene Therapy Development Success

AI in cell and gene therapy is now a necessity for successful development. As pipelines expand, sponsors are turning to machine learning to solve problems that conventional methods simply couldn’t handle at scale.

Image

Why Cell and Gene Therapy Needs AI Now

Artificial intelligence (AI) is rapidly becoming a critical component of cell and gene therapy (CGT) development.

While early discussions focused on AI-assisted drug discovery, the technology is now influencing every stage of the CGT lifecycle, from target identification and vector design to manufacturing, quality control, and regulatory decision-making.

As cell and gene therapies move towards commercial scale, AI is emerging as more than a tool for automation but as a strategic platform for overcoming some of the sector's biggest scientific and operational challenges.

According to a recent review in Pharmaceutics, over 3,200 active clinical trials were investigating gene, cell, or RNA therapies globally by late 2025, with the gene therapy market projected to grow from roughly ten billion US dollars in 2025 to as much as 118 billion US dollars by 2035.

That growth collides with four problems that traditional pharma methods struggle to solve:

• Biological complexity: Living cells and gene constructs behave unpredictably, so outcomes are hard to model in advance
• Manufacturing intricacy: Personalised, autologous therapies like CAR-T require a patient-specific, ‘vein-to-vein’ supply chain
• Scale-out, not scale-up: CGT manufacturing runs thousands of small parallel batches rather than one large batch
• Translational risk: Results in animal models often fail to predict what happens in human patients

Read Maximise Pharma Innovation With Quality-by-Design

Maximise Pharma Innovation With Quality-by-Design

Find out how Quality-by-Design improves reliability and compliance in pharma and how you can maximise innovation through quality-based pharma development systems.

Cell and Gene Therapies Are Transforming Medicine, but Development Remains Difficult

Cell and gene therapies have reshaped expectations for treating diseases that were previously considered untreatable.

Gene editing, viral vectors, CAR T-cell therapies, stem cell therapies, and RNA-based medicines are creating new opportunities across oncology, rare diseases, inherited disorders, and regenerative medicine.

However, scientific breakthroughs alone do not guarantee commercial success.

Compared with conventional pharmaceuticals, CGTs present a far more complex development pathway.

Manufacturing is highly variable, biological materials are inherently difficult to standardise, and small process changes can significantly affect product quality, potency, and safety.

But now AI is increasingly helping CGT developers address these challenges by analysing complex biological data, improving manufacturing consistency, supporting quality-by-design (QbD), and enabling faster decision-making across development.

Rather than replacing scientific expertise, AI functions as an intelligent decision-support system that strengthens both research and technical operations.

AI in Cell and Gene Therapy Development: Construct Design

Designing vectors and CAR constructs faster

Conventional construct design relies on building and testing large libraries of variants by hand. Generative AI flips this process.

Machine learning models can now predict how a viral vector's capsid sequence affects tropism and immunogenicity, before a single molecule is synthesised.

The same approach helps optimise the binding domains of CAR-T receptors, balancing strong target binding against the risk of unwanted, low-level signalling.

Image

Simulating patient response with digital twins

Digital twins, virtual models built from a patient's genomic, imaging, and clinical data, let developers simulate how a therapy might behave before it reaches a person.

These models can forecast the persistence of CAR-T cells in the body and flag the risk of serious side effects such as cytokine release syndrome, allowing clinicians to intervene earlier.

Viral vector engineering

Efficient delivery remains one of the biggest challenges in gene therapy.

Adeno-associated viruses (AAVs), lentiviral vectors, and other delivery platforms must balance tissue specificity, transduction efficiency, manufacturability, and safety.

AI helps researchers optimise vector engineering by predicting how modifications to viral capsids influence biological performance.

Instead of relying solely on trial-and-error experimentation, computational models can evaluate thousands of potential sequence variations before laboratory validation begins.

This shortens optimisation cycles while reducing development costs.

Several biotechnology companies are already applying machine learning to engineer next-generation viral vectors with improved tissue targeting and reduced immunogenicity.

Optimising manufacturing processes

Perhaps the greatest opportunity lies within manufacturing.

Unlike traditional small-molecule medicines, every batch of a CGT contains living biological material.

Maintaining consistency throughout manufacturing remains one of the industry's most significant technical challenges.

There are several areas where AI is already improving production.

These include:

• optimisation of cell culture conditions
• prediction of manufacturing outcomes
• bioreactor process control
• media optimisation
• automated process adjustments
• predictive maintenance
• yield optimisation

Rather than responding after deviations occur, AI allows manufacturers to identify subtle process changes before they affect product quality.

This shift from reactive quality management towards predictive process control aligns closely with modern pharmaceutical manufacturing strategies.

Read The Looming Threat of Geopolitical Risk in Clinical Supply Chains

The Looming Threat of Geopolitical Risk in Clinical Supply Chains

Geopolitical risk is reshaping how clinical supply chains are managed, demanding resilience and foresight. Learn why building supply chain resilience is the key for clinical trial success.

Image

Smart Manufacturing on the Production Floor

Manufacturing is where CGT's therapies most often meet their biggest bottleneck.

Combatting this, a digital twin of a bioreactor run can test thousands of virtual scenarios before a physical batch is even started, flagging the ideal temperature, nutrient feed, and timing needed to hit quality targets.

Once manufacturing begins, AI systems track critical process parameters, such as glucose levels and cell viability, in real time. If a batch begins to drift off course, the system can send commands back to the equipment automatically, catching problems before an entire batch is lost.

The same digital infrastructure tracks a therapy's location, temperature, and integrity across its full journey from patient to lab and back again.

Digital manufacturing is moving towards real-time decision-making

Another major theme throughout the review is the emergence of digitally connected manufacturing facilities.

Cell and gene therapy manufacturing already generates enormous quantities of process data from sensors, analytical instruments, imaging systems, and laboratory information platforms.

Historically, much of this information has been analysed retrospectively. AI changes that.

Real-time analytics allow manufacturers to detect emerging process variation while production is still underway.

Digital twins, predictive analytics, and intelligent automation can simulate manufacturing performance before changes are implemented, reducing risk while supporting continuous process improvement.

These capabilities will become increasingly important as commercial manufacturing expands and production networks become more distributed globally.

Read Cell Therapy for Solid Tumours: Current Challenges and New Advances

Cell Therapy for Solid Tumours: Current Challenges and New Advances

Cell therapy for solid tumours faces unique biological barriers. This analysis unpacks the engineering strategies pharma leaders must understand for R&D success in 2026.

AI Is Becoming Central to Quality-by-Design

Quality assurance has traditionally relied on end-product testing.

However, regulators increasingly encourage manufacturers to build quality into the manufacturing process itself.

Now AI can support Quality-by-Design (QbD) by continuously analysing manufacturing data throughout production.

Image

Machine learning models can identify relationships between critical process parameters (CPPs) and critical quality attributes (CQAs), helping manufacturers understand how process variation influences product performance.

Examples include predicting:

• cell viability
• viral vector yield
• Potency
• impurity profiles
• process deviations

Instead of generating static manufacturing protocols, AI enables continuously improving production systems that learn from previous manufacturing runs.

For developers seeking commercial scalability, this represents a significant competitive advantage.

Regulation and Safety Monitoring Are Catching Up

Regulators are adapting existing frameworks rather than writing CGT-specific AI rules.

The FDA's FRAME initiative supports the adoption of AI-driven manufacturing technology, while separate 2025 draft guidance addresses how AI-derived data should be presented for regulatory review.

The EMA's 2024 reflection paper sets out similar expectations for data governance and transparency across the European Union.

Regulators understand that safety monitoring benefits too. Natural language processing can now scan adverse event reports and unstructured clinical notes automatically, while machine learning links subtle manufacturing deviations, even ones from years earlier, to safety signals that emerge much later in a patient's life.

Read Designing Decentralised Clinical Trials for Patient Diversity

Designing Decentralised Clinical Trials for Patient Diversity

Data shows that most DCT approaches significantly improve patient diversity. This analysis examines what the evidence means for clinical trial design strategy.

Image

There Are Still Gaps in AI in CGT

The picture is not all rosey.

Data scarcity remains the biggest limitation: CGT patient populations are small, data sits in silos, and rare diseases mean thin datasets by default.

Direct, head-to-head comparisons between AI-driven methods and traditional design-of-experiments approaches are also still rare, so the incremental benefit of AI is not always easy to quantify.

Algorithmic bias and the “black box” nature of deep learning models add a further layer of scrutiny for regulators and clinicians alike.

Data quality

Machine learning models require large, high-quality datasets.

However, many CGT programmes involve relatively small patient populations, limited manufacturing batches, and highly specialised datasets.

Generating sufficient high-quality training data remains a significant challenge.

Standardisation

Different organisations often collect manufacturing and analytical data using different protocols, platforms, and formats.

Improved data harmonisation will be essential if AI models are to be applied consistently across organisations and manufacturing sites.

Regulatory confidence

As AI increasingly influences manufacturing decisions, regulators will expect developers to demonstrate transparency, validation, reproducibility, and robust governance.

Explainable AI will become increasingly important for regulatory acceptance, particularly where algorithms contribute to critical manufacturing or quality decisions.

Workforce capability

Successful implementation depends on more than software.

Organisations require multidisciplinary teams capable of combining expertise in biology, engineering, manufacturing, computational science, data analytics, and regulatory affairs.

Developing this workforce will be just as important as investing in AI platforms themselves.

AI Will Define the Next Generation of CGT Manufacturing

AI in CGT is now a foundational capability that supports research, development, manufacturing, quality assurance, and commercial scale-up.

Importantly, the greatest value does not come from replacing existing scientific processes.

Instead, AI strengthens human decision-making by integrating complex biological, manufacturing, and analytical data into actionable insights.

As commercial demand for cell and gene therapies continues to grow, manufacturers will need to increase productivity while maintaining exceptionally high quality standards.

AI provides one of the most promising pathways towards achieving that balance.

For pharmaceutical executives, the strategic question is therefore shifting. Success will depend less on whether AI is adopted and more on how effectively organisations integrate AI across the entire CGT value chain, from discovery through commercial manufacturing.

Companies that build data-driven development ecosystems today are likely to be better positioned to accelerate innovation, improve manufacturing robustness, and bring advanced therapies to patients more efficiently.

At Pharmatica, we analyse the technologies transforming pharmaceutical innovation, connecting scientific advances with the strategic decisions shaping the future of drug development. From AI-enabled discovery to next-generation manufacturing, our expert Insights help life sciences leaders navigate an increasingly data-driven industry.

Pharmatica: Insight. Connection. Impact.

Image