Integrating AI and machine learning in drug discovery and development

Pharmaceutical companies face a variety of challenges when researching and developing potential new products, including the “patent cliff” and the difficulty of discovering new “blockbuster” drugs (1). Pharmaceutical companies are integrating AI and machine learning (ML) as part of their drug development toolbox, not just during discovery and development, but throughout the drug lifecycle (1, 2). This technological shift is not just about automation, but rethinking how treatments are discovered, tested, and brought to market (1, 3).

A report from Capgemini Research Institute details how the industry is moving to AI to reduce costs and clinical failure rates (4). In a survey of 500 senior executives from eight countries, a majority (82%) believe that AI will fundamentally change biopharmaceutical research and development, and 63% say that failing to scale AI will leave companies behind in terms of innovation and market relevance (4). They also predict that over the next decade, most new molecular entities will be created using AI-driven platforms.

How is AI accelerating early-stage discoveries and solutions?

Low solubility of new chemicals is a constant challenge in drug development, with an estimated 70–90% of drug candidates classified as poorly soluble (5, 6). Traditionally, identifying the appropriate formulation of these “brick dust” or “greaseball” molecules has required extensive and expensive empirical trial and error (5,6). but, in silico modeling The use of AI and ML is currently being leveraged to predict physicochemical properties and guide early-stage decision-making (5,6). Researchers can utilize tools such as quantum mechanical calculations and molecular dynamics simulations to identify optimal solubilization methods and appropriate excipients (5). This “super material-sparing” approach significantly reduces API consumption and saves research costs (5). A notable example is the evaluation of compound CVN424, where predictive modeling guided the selection of the spray-dried formulation and accelerated the overall development schedule (6).

How are digital twins improving personalized clinical assessments?

The use of digital twins in preclinical evaluation is revolutionizing the way drugs are tested in human organs (7). for example, in vitro Lung perfusion systems allow researchers to keep human lungs alive outside the body and collect “clean” multimodal data. A machine learning model then uses these data to create a personalized digital control arm for each organ being treated.

This approach makes it possible to directly compare observed treatment effects with “untreated” results generated by digital twins within the same organ. By reducing reliance on large control groups and traditional animal models, digital twins have the potential to reduce research scale and accelerate clinical trials (7).

How are regulatory workflows and market access streamlined?

New European Union regulations requiring complex collaborative clinical assessments are increasing the burden of regulatory compliance and health technology assessment. Generative AI (GenAI) is being utilized to manipulate raw data to complete these submissions by automating the creation of clinical evidence summaries (8), reducing initial documentation time by approximately 40% in some cases.

Beyond documentation, agent AI can automate the labor-intensive process in clinical trial master file management of capturing and classifying documents with “human-involved” checkpoints for quality control (9).

Domain-specific models and secure collaboration

Although general-purpose AI models are proficient in language, they often lack the depth needed for highly regulated pharmaceutical markets (9). To address this, domain-specific language models can be trained based on unique company documentation, such as batch records and validated procedures that support high-stakes decisions regarding drug manufacturing and clinical protocols (10).

To improve communication between regulators and drug developers, secure cloud environments like the PRISM project allow both parties to work on the same document in real time and use features like AI agents to accelerate the regulatory process and ensure safety and efficacy are addressed long before a product reaches a patient (1, 2).

What AI tools are entering the drug development space?

In recent years, drug developers have invested significant time, money, and effort into oncology. In January 2026, SOPHiA GENETICS, an AI-powered precision medicine company, announced a partnership with MD Anderson Cancer Center that leverages SOPHiA GENETIC’s AI-powered analytics platform. The two organizations will launch a research and development program to jointly develop next-generation sequencing oncology tests that translate complex multimodal data (11).

Also in January, AI drug discovery company Oxford Drug Design announced the successful completion of this study. in vivo Validation of novel therapeutic approaches across multiple tumor types in the development of potential first-in-class cancer therapies using the GenAI platform. “In studies using genetically engineered mouse models that recapitulate the earliest mutational events in colorectal cancer, Oxford Drug Design’s lead compound demonstrated statistically significant anti-tumor activity with efficacy comparable to the benchmark therapy rapamycin, with no detectable signs of toxicity,” the company explained in a press release (12).

Iktos, a company applying AI and robotics to drug discovery, has signed a multi-target collaboration agreement with pharmaceutical company Servier to leverage Iktos’ AI-tuned discovery platform to accelerate the design and optimization of new small molecule therapeutics in oncology and neurology. Iktos applies generative AI and AI-tuned robotics platforms to design, synthesize, and optimize small molecules for multiple undisclosed targets. They may then be evaluated and selected by Servier for preclinical and clinical development (13).

The human element in the digital ecosystem

Experts emphasize that AI and other digital technologies are power multipliers, not replacements for human expertise (7,9). Successful integration requires a “digitally proficient” workforce that can oversee automated systems and interpret impact assessments generated by AI (3). AI automates administrative burdens, allowing scientists and regulatory professionals to return to the core of their work: scientific reasoning and clinical judgment (2, 8).

Responsible adoption of AI and machine learning creates more efficient and equitable paths to innovation, accelerating the time from molecular discovery to patient treatment (8).

References

1. Cole, C. Global collaboration and technology acceleration to advance pharmaceutical innovation. PharmTech.comDecember 24, 2025.
2. Cole, C. Leveraging data and a secure cloud environment for global pharmaceutical solutions. PharmTech.comDecember 23, 2025.
3. Cole, C. Organizing the future of pharma: Integrating living decision engines and AI. PharmTech.comDecember 23, 2025.
4. capgemini. Smart bet, option only, or both?: Biopharmaceutical R&D focuses on AI. https://www.capgemini.com/insights/research-library/how-ai-will-transform-life-sciences-r-and-d/ (Accessed 13 January 2026).
5. Mirasol, F. Lavery, P. Leveraging AI/ML to reduce risk in drug development: Part 1. PharmTech.comNovember 18, 2025.
6. Cole, C. Overcoming low solubility challenges using AI and molecular dynamics. PharmTech.comNovember 10, 2025.
7. Cole, C. Mirasol, F. Enabling a personalized digital control arm for preclinical drug evaluation. PharmTech.comNovember 13, 2025.
8. Dhas, A. Accelerating HTA preparation with generative AI. PharmTech.comDecember 17, 2025.
9. Medable. Medable launches new AI agent to automate clinical trial document management, speaks at JPM Health. press release. January 6, 2026.
10. Nair, J. Why pharmaceutical company AI agents need small domain-specific models first. pharmaceutical technologyDecember 3, 2025.
11. Sophia Genetics. SOPHiA GENETICS and MD Anderson announce a strategic partnership to accelerate AI-driven precision oncology. press release. January 7, 2026.
12. Oxford drag design. Oxford Drug Design makes further announcements in vivo Verification of novel tumor treatment mechanisms. press release. January 12, 2026.
13. Ictos. Iktos enters a multi-target AI strategic partnership worth over €1 billion with Servier to advance drug discovery in oncology and neurology. press release. January 8, 2026.

About the author

Susan Hagney is pharmaceutical technology®.