HealthManagement, Volume 24/25 - Issue 6, 2025

img PRINT OPTIMISED
img SCREEN OPTIMISED

AI is revolutionising the biopharma industry by transforming drug discovery, development and manufacturing processes. It accelerates drug discovery, reduces costs and improves precision in disease modelling and drug design. AI enhances workflow efficiency, optimises manufacturing and refines supply chains while supporting vaccine design and precision medicine. However, challenges like regulatory transparency and data biases remain critical for broader adoption across the sector.

 

Key Points

  • Drug discovery can be accelerated by reducing timelines and costs with the help of AI.
  • Better precision in disease modelling and drug design can be achieved with AI.
  • AI improves manufacturing efficiency and supply chain optimisation.
  • AI aids vaccine design and precision medicine advancements.
  • The adoption of AI faces challenges with data bias and regulatory transparency.

 

Artificial Intelligence (AI) is reshaping the way the biopharma industry functions, transforming how drugs are discovered, developed, manufactured and delivered. From shrinking drug discovery timelines to optimising biomanufacturing processes and making supply chains more transparent and resilient, AI has shown promise in closing the translational gap from bench to clinic. Its transformative impact across the biopharma value chain has witnessed steady growth, driven by advances in machine learning algorithms, computational power and the availability of large datasets. Newer AI platforms are extensively working towards eliminating the gaps in data-driven drug development, clinical trials, bioprocessing and supply chain to make it an integral part of the biopharma value chain. Everest Group is a global research firm guiding business leaders with tailored strategies for operational and financial success through expert insights in technology, business processes and engineering. Its advanced SciTech service line (AST) provides actionable research on cutting-edge science and technology innovations, focusing on R&D trends, innovation processes, toolkits and critical future drivers.

 

Drug discovery and development is notoriously complex and daunting, characterised by long timelines, exorbitant costs and high attrition rates. While the most apparent impacts on cost reduction and compressing timelines remain the most significant advantages, its use in accurate disease modelling, novel target discovery and de novo drug design to design drugs with desired drug-like properties is garnering traction. The following figure depicts the evolving role of AI in biopharma applications.

 

 

AI for Enhanced Workflow Efficiencies

With its advanced learning models to analyse large amounts of complex and disparate data in crunched timelines, AI makes its mark in drug discovery, disease modelling, target discovery and precision medicine. Large pharma and biopharma companies are establishing themselves as pioneers in using AI in drug discovery, either with internal programmes or collaborating with AI-driven drug discovery platform developers to identify novel targets and hit-to-lead identification. Recent examples include Healx, which will use its AI-based drug discovery technology to analyse proprietary Sanofi compound data and identify potential rare disease targets. At the same time, integrating AI with in vitro disease models like organ on chips and organoids can bring out nuances and guide precision medicine strategies.

 

AI has been heavily exploited for its role in improving understanding of PPIs, protein-ligand interactions and protein folding. In 2024, Google DeepMind’s latest launch, AlphaFold3, has further expanded the scope of AI to predict protein folding and interactions accurately. It is now an open-source model that is being used by thousands of scientists globally.

 

The use of multimodal AI is becoming increasingly crucial across the pharmaceutical value chain. It provides unprecedented insights into hidden linkages in pathways and helps understand disease progression that human eyes might miss. The integration of diverse datasets (omics, imaging, clinical outcomes) reveals deep connections and helps understand from a more holistic perspective. Sophia Genetics and Astra Zeneca have recently collaborated to use Multimodal AI for precision oncology (lung cancer patient stratification).

 

AI is also useful for designing synthetic pathways that minimise the number of steps and chemicals required and increase yields. Elsevier and Iktos recently collaborated to use AI-driven synthetic chemistry automation to accelerate the Design-Make-Test cycle in drug discovery.

 

While most solutions address specific aspects such as virtual screening or lead optimisation, companies are also developing integrated solutions that can be used across the drug discovery and development continuum. One such solution is Merck’s Addison software, launched in December 2023. It is the first AI solution to integrate discovery and synthesis in a single platform. It uses generative AI to identify the most suitable drug-like candidates from a vast chemical space of 60 billion options and to provide an optimal synthesis route.

 

De novo synthesis is one of the most exciting frontiers of AI, and generative AI models are being used to design both small molecules and biologics, with the latter being at a more nascent stage. The use of GAN (generative adversarial networks) in Generative AI in drug discovery workflows is expected to grow steadily and be most impactful, though developments are still nascent. Qure.ai, Niramai Health, Owkin and InSilico Medicine are some companies advancing GenAI to create large quantities of synthetic data for de novo drug design, novel drug target discovery, design precision clinical trials and prediction of drug responses, which will have a radical impact. InSilico Medicine is one of the pioneers in this space, developing one of the world’s first GenAI-developed drugs for Idiopathic Pulmonary Fibrosis that entered Phase 2 trials in 2023.

 

Broadening Scope of AI Across Modalities

While AI is broadly adopted and used for small molecule discovery and design, its application in the design of vaccines, biologics and advanced therapeutic modalities is progressing considerably. Sanofi and other companies have used generative AI to accelerate the design of new mRNA vaccines and to optimise their delivery and formulation. Leveraging AI in vaccine design and development is also promising, as it can accelerate immunogen design and antigen selection as well as predict immune responses. Sanofi, Moderna, GSK and several other biopharma companies have adopted AI to accelerate mRNA vaccine design and are partnering with AI platform companies to streamline their discovery efforts.

 

Evolving role of LLMs

The integration of Large Language Models (LLMs) into drug discovery and development is unlocking new possibilities across the entire value chain. LLMs represent a significant opportunity for companies that goes beyond discovery. AI-powered LLMs can analyse vast amounts of unstructured data and fine hidden patterns from large datasets for data integration and triangulation, providing predictive insights.

 

These platforms have also been democratised, opening up opportunities for pharmaceutical companies to develop innovative solutions, improve drug discovery, clinical operations, commercialisation, clinical trials and, finally, therapeutic outcomes. Insilico Medicine is using Microsoft’s BioGPT, a LLM model, to predict dual-purpose targets, while ConcertAI and NIH’s National Library of Medicine (NLM) are deploying a platform called TrialGPT for rapid patient selection and recruitment in clinical trials. Moderna has also collaborated with OpenAI to utilise its ChatGPT platform to accelerate mRNA vaccine development. Google’s recent launch of the drug discovery-specific LLM platform, TxLLM, is a refined version of MedPALM2, which can predict interaction and screen compounds and is expected to gain traction. The following figure represents the role of AI in drug discovery and development.

 

Expanding Role of AI Beyond Drug Discovery and Development

AI is revamping the pharma manufacturing and innovation process with its ability to enhance efficiency, drive precision and reduce time across the drug development and manufacturing lifecycle. Its applications include such options as:

 

Process Optimisation and Control. AI algorithms can analyse complex datasets to optimise manufacturing processes, ensuring consistent product quality and reducing production costs. LLM and GenAI models can help accurately predict process parameters, enabling real-time adjustments, minimising variability and optimising production processes. AI-driven advanced process control (APC) systems are being explored to dynamically manage pharma manufacturing to ensure product consistency and achieve desired outputs.

 

Predictive Maintenance and Quality Control. One of the key advantages of AI implementation in pharma manufacturing is predictive maintenance, which helps anticipate risks and plan to overcome potential pitfalls. AI-integrated sensors can prevent unexpected downtime and extend equipment lifespan. AI models can accurately detect anomalies and deviations in real time and identify patterns that can lead to potential quality issues, allowing for immediate corrective actions. This is a boon for an industry governed by stringent regulatory standards. AI-driven digital twins enable companies to predict how changes in process parameters affect product quality and yield, facilitating continuous improvement and innovation.

 

Supply Chain Optimisation. One of AI's key advantages is its ability to optimise supply chain management by predicting demand, managing inventory and identifying potential disruptions. Machine learning algorithms can analyse market trends, historical data and external factors to forecast demand and plan production schedules. They can also track the movement and distribution of drugs and other products, ensuring transparency and preventing counterfeiting.

 

With proven advantages and implementation use cases, several pharmaceutical players are actively integrating AI into their manufacturing and supply chain processes. The application segment has also seen companies collaborate to form innovation networks and encourage the start-up ecosystem. One example is AION Labs, an Israeli venture studio focused on adapting AI and machine learning in the pharmaceutical industry. With Tier 1 companies such as AstraZeneca, Merck KGaA, Pfizer and Teva Pharmaceuticals being active participants, AION Labs has ventures like DenovAI, which utilises AI for antibody discovery.

 

Reducing AI Black Boxes to Navigate Regulatory Pathways

The bias surrounding the use of AI tools greatly limits their value and reliability. To keep pace with AI advances, regulatory bodies need to develop regulatory frameworks to guide the application of AI in the biopharma sector. Several stakeholders, including platform developers and pharma companies, are focused on determining how to make AI tools more precise and predictable.

 

Developing algorithms using Explainable AI, which can break into the “AI black boxes” and provide a rationale for AI algorithm decisions and outcomes. These transparent models can meet regulatory standards and help build stakeholders' trust. Building models based on biological mechanisms, which consider pathways and interactions, can be useful to align AI predictions. Regulatory bodies can use such transparent reasoning for predictions and recommendations for AI-developed drugs.

 

The transformative potential of AI and other digital technologies is also recognised by governmental organisations that actively participate in creating regulations and guidelines for AI use in the pharma industry, facilitating its integration into the sector. In May 2023, the FDA initiated two discussion papers on Artificial Intelligence and Machine Learning in Drug Discovery and Manufacturing, inviting industry feedback to understand the necessary elements for implementing AI-based models in a cGMP environment. This aims to establish guidelines for AI's safe and effective use in pharmaceutical manufacturing. Similarly, the European Union (EU) is also developing regulations to ensure AI's safe, reliable and ethical use.

 

Future Outlook

The biopharmaceutical industry is at a transformative phase, driven by the convergence of digital technologies with traditional innovations. The critical need for drug development, manufacturing and patient care innovation offers unprecedented opportunities for AI to accelerate discovery, reduce costs and personalise treatments, making it a pivotal tool for addressing the industry’s challenges. Its ability to democratise healthcare by acting as a lever for accessible care, globalising health applications and optimising resources makes its role inevitable in the future of the healthcare industry.

 

Its adoption in healthcare, especially in biopharma applications, depends on ensuring it navigates the regulatory and ethical challenges. While regulatory bodies are receptive to AI tools, they demand transparency and validation. Consequently, it is the responsibility of stakeholders and technology developers to ensure that AI models are:

 

  • Explainable: ensuring AI models are interpretable and explainable for regulatory approval;
  • Adhering to data privacy: implementing secure measures to protect patients’ or stakeholders’ data privacy;
  • Mitigating bias-related issues: addressing datasets' biases to ensure equitable outcomes.

 

Addressing the above issues requires a collaborative approach from various stakeholders to strengthen the data infrastructure and build AI expertise within all walks of the industry. This includes building robust AI models and standardising data infrastructure, fostering collaboration and enhancing data quality. It also requires upskilling the workforce and fostering multidisciplinary collaborations. In order to facilitate AI adoption in the industry, it is necessary to address ethical concerns, which include standardising reporting procedures, mandating ethical frameworks and employing data privacy measures, such as advanced encryption and de-identification methods. It is also important to understand that the use of AI tools is an iterative process that requires continuous monitoring, evaluation and conducting of pre- and post-deployment audits, as well as establishing and implementing feedback loops.

 

In conclusion, while the future of AI in the biopharma industry is exciting, it is also complex. To fully exploit AI's potential, the sector must take proactive steps to address technical, ethical and regulatory challenges.

 

Conflict of Interest

None

 


References:

Chun M (2023) How Artificial Intelligence is Revolutionizing Drug Discovery. Bill of Health – The Blog of the Petrie-Flom Center at Harvard Law School [Accessed: 9 January 2025] Available from blog.petrieflom.law.harvard.edu/2023/03/20/how-artificial-intelligence-is-revolutionizing-drug-discovery/

Claritax Rx (2024) AI in Biopharma: Uncovering the Latest Game-Changers in Drug Discovery. [Accessed: 9 January 2025] Available from claritasrx.com/blog/ai-in-biopharma/

Hagen J (2024) Google creates Tx-LLM for drug discovery and therapeutic development. MobiHealthNews [Accessed: 9 January 2025] Available from mobihealthnews.com/news/google-creates-tx-llm-drug-discovery-and-therapeutic-development

Kint S, Dolfsma W, Robinson D (2024) Strategic partnerships for AI-driven drug discovery: The role of relational dynamics. Drug Discovery Today, 29(12): 104242 [Accessed: 9 January 2025] Available from sciencedirect.com/science/article/pii/S1359644624003672

Mirakhori F, Niazi SK (2025) Harnessing the AI/ML in Drug and Biological Products Discovery and Development: The Regulatory Perspective. Pharmaceuticals, 18(1), 47 [Accessed: 9 January 2025] Available from mdpi.com/1424-8247/18/1/47

O’Shea C (2024) Revolutionising biomedical data sharing and AI-driven healthcare solutions. Trinity College Dublin [Accessed: 9 January 2025] Available from tcd.ie/news_events/articles/2024/-revolutionising-biomedical-data-sharing-and-ai-driven-healthcare-solutions/

Oualikene-Gonin W, Jaulent MC, Thierry JP et al. (2024) Artificial intelligence integration in the drug lifecycle and in regulatory science: policy implications, challenges and opportunities. Front Pharmacol, 2(15): 1437167 [Accessed: 9 January 2025] Available from pmc.ncbi.nlm.nih.gov/articles/PMC11327028/

PR Newswire (2023) Merck Launches First Ever AI Solution to Integrate Drug Discovery and Synthesis [Accessed: 9 January 2025] Available from prnewswire.com/in/news-releases/merck-launches-first-ever-ai-solution-to-integrate-drug-discovery-and-synthesis-302003351.html

Satterlee K (2024) The Next Frontier for Vaccines: Using AI to Predict Antigens. Texas A&M University College of Engineering [Accessed: 9 January 2025] Available from engineering.tamu.edu/news/2024/09/the-next-frontier-for-vaccines-using-ai-to-predict-antigens.html

Walrath R (2025) Pfizer doubles down on AI partnership with antibody-drug conjugate deal. Chemical & Engineering News [Accessed: 9 January 2025] Available from cen.acs.org/business/start-ups/Pfizer-doubles-down-AI-partnership/103/web/2025/01