GenAI's Impact on Reshaping the Pharma Industry
SAMEER LAL, SVP, Enterprise Medical Solutions, Indegene
JOHN WARD, Director, Ask GXP & ServBlock
ANDREE BATES, Expert, Artificial Intelligence (AI), Pharmaceuticals, and Strategic Implementation of AI
Sameer Lal: In an interview, Sameer Lal discusses the impact of GenAI in reshaping the pharma industry. Exploring GenAI's transformative potential, emerging trends, and regulatory challenges, this discussion offers valuable insights for pharmaceutical executives navigating the evolving landscape of AI in drug development and commercialisation. It also sheds light on the potential impact of GenAI adoption on the pharma workforce. John Ward: The integration of Generative AI (GenAI) technologies in the pharmaceutical industry represents a significant paradigm shift, promising to enhance drug discovery, streamline manufacturing processes, and optimise knowledge management. We discuss the methodology of implementing GenAI tools, such as automated document generation, AI-driven simulation models for clinical trials, and adaptive learning systems for workforce training. Andree Bates: Generative AI (GenAI) holds tremendous promise for automating drug design, speeding up clinical trials with synthetic data and digital twins, automating many aspects of regulatory and medical affairs and personalising sales and marketing. Nevertheless, fully realising these advancements while managing associated risks demands meticulous planning from both industry and regulators.
How is AI and especially generative AI (GenAI) starting to transform the pharma industry? What advantages does it offer over traditional methods, and what are the limitations?
SAMEER: From my vantage point, over the past year, the initial hype as well as the wariness around GenAI has started to settle into cautious optimism. Within the confines of their secure platforms, pharma enterprises have been experimenting with use cases right from R&D to medical affairs to commercial functions. GenAI specifically is allowing a fundamental re-think on three broad fronts: summarising and generating insights from vast amounts of data; generation of content; and healthcare professional (HPC) engagement. Each of these in turn is being used by pharma to serve any (or a mix) of these objectives—efficiency, effectiveness, or strengthening competitive edge.
Having said that, in the absence of clear regulations, data security and privacy concerns, clarity over copyrights as well as overall risk-averseness of the industry implies that scaling up from these experiments will take some time to come to fruition.
JOHN WARD: The introduction of GenAI in pharmaceutical manufacturing is shifting the landscape for knowledge workers. By automating routine documentation, data entry, and monitoring tasks, workers are freed up to focus on more strategic and creative tasks such as process optimisation and innovation. This shift not only enhances productivity but also improves job satisfaction by allowing workers to engage in more meaningful and rewarding work.
ANDREE BATES: Generative AI models possess the potential to revolutionise drug discovery, research and development, streamline clinical trials, improve patient care, and enhance numerous aspects of regulatory affairs, medical affairs and personalise sales and marketing at scale. Examples of gen AI applications in Pharma include:
- Creating New Drugs/Drug Discovery: Generative AI can make new molecules that are specifically designed to deal with certain diseases or health problems. The molecules can be made to work better and be safer. They can also be created so they can be absorbed, distributed, metabolised, and excreted more effectively in the body.
- Improving Drug Candidates/Lead Optimisation: Generative AI can help with this by making new molecules that are similar to the drug candidate but have better properties. These new molecules can then be made and tested to see how well they work and how safe they are.
- Finding New Uses for Drugs/Drug Repurposing: Another use for Generative AI is to find new uses for drugs that already exist. Generative AI can do this by making new molecules that are similar to existing drugs but work differently. These new molecules can then be tested to see if they could be used to treat other diseases.
- Creating Personalised Medicines: Generative AI can help create these personalised drugs by making new molecules that are specifically designed for certain groups of patients. It can help make treatments work better and have fewer side-effects by making new molecules that are specifically tailored to individual patients.
- Clinical Trials: Generative AI can make clinical trials more efficient by figuring out which groups of patients are likely to respond well to a certain treatment. For example, generative AI programmes are being trained to find genetic markers that can predict if a patient will respond well to a certain drug. It can also speed up trials by finding the patients using AI, and leveraging Gen AI-generated synthetic data and digital twins reducing the need for as many humans in the trials.
- Regulatory: Automated document sorting, text recognition, automated filling of standard regulatory forms, automatically classify and tag various types of documents, ensuring that every submission is complete and correct, collating daily guidance, regulations, news and Predict compliance changes.
- Medical Affairs: Intelligent literature monitoring, track and respond to large volumes of medical queries, Automated medico-legal reviews, automated SLRs, automated literature monitoring.
- Sales and Marketing: Create content and assist with personalised interactions (Websites, social media, and messaging), and create personalised messages (emails and marketing materials), precisely target the right physicians, identify patients who will stop adhering in advance and much more.
The advantages over traditional methods are the ability to analyse extensive datasets rapidly, formulate predictions, and generate new data and insights making them indispensable tools for researchers, clinicians, and pharmaceutical teams in general. In particular, some of the Generative AI models being employed for this are Generative Adversarial Networks (GAN), Recurrent neural networks, Variational auto-encoders, Deep Reinforcement learning and Transformer models.
What emerging trends do you foresee in the intersection of GenAI and pharma? How might advancements in AI technologies reshape the industry in the coming years?
SAMEER: We are seeing some excellent results come through in all the 3 areas that I mentioned earlier. Here are some specific examples:
- Summarisation and generation of insights: A clear use-case is in reviewing and analysing vast amounts of information that gets presented at conferences like ASCO. GenAI allows the generation of quick summaries from large articles and actionable insights to be curated from there
- Content generation is likely to face the most change in the coming years. From “writing” of the first draft of various regulatory and safety documents, creation of plain language summaries, translation from one language to another, and transcreation from one format to another (Word to infographic) all have been successfully demonstrated. With Sora, the creation of video from text is now child’s play and does not need a full-service agency
- Customer engagement is already turning on its head with conversational AI implemented in chatbots, speech-to-text conversion in call centres, as well as hyper-personalisation possibilities that hitherto were not even possible
The pharma industry which has always been at the forefront of innovation will now need to assess their use of the vast data that at its disposal to create small LLMs for their own use, reengineer process flows, and rethink traditional roles and job descriptions to get the most value from this change.
JOHN WARD: Revolutionising Knowledge Management: GenAI excels in organising and processing large volumes of data, transforming how knowledge is managed on the shop floor. It can analyse historical data to identify trends and patterns, suggest process improvements, and even predict maintenance needs before breakdowns occur. This proactive approach to maintenance and operations not only enhances efficiency but also extends the lifespan of manufacturing equipment.
ANDREE BATES: The industry has the potential to be streamlined with Gen AI taking on more roles within identifying new molecules, speeding up clinical trials, and increasing efficiency throughout most functions. With the advent of whole-body digital twins, we are likely to see pharma stop doing manufacturing and each patient will have personalised dosages of medicines as determined by their digital twin. Each pharmacy will have a 3D printer and print the drug for the patient. 3D printers that already print biologics exist so this is not as far-fetched as you may think!
How do you envision Gen AI shaping the future of clinical trials, regulatory submissions, and drug commercialisation?
SAMEER: I anticipate that the traditional timelines for a drug to reach from lab to market are likely to get shortened as GenAI makes its impact felt through the drug development value chain. Starting from the lab through high-velocity screening of drug candidates to clinical trials where patient recruitment can be upended to generating regulatory documents for submission to auto-generate personalised content for rep/MSL interaction with HCP—each stage will be impacted with GenAI.
JOHN WARD: Streamlining Document Generation and Management: GenAI tools can automate the creation, retrieval, and storage of documents. They ensure that the latest versions of documents are easily accessible, reducing the time workers spend searching for information. This capability is crucial in highly regulated environments where compliance and up-to-date documentation are paramount.
ANDREE BATES: Gen AI will reshape clinical trials in many ways such as faster patient enrolment, synthetic data control arms and digital twin studies, automated data analysis and automated CSR writing. Regulatory submissions will be largely automated and the time from bench to market will be sped up by between 60-70%.
How is the adoption of GenAI affecting the pharma workforce? What new skill sets are required, and how can companies facilitate reskilling and upskilling initiatives?
SAMEER: I do not anticipate the role of humans to ever disappear once GenAI starts getting implemented at a scale that will justify the promise. But the roles will definitely evolve. Let me give you a specific example. Take the case of medical writers; with the use of Gen AI, writers will become subject matter experts (SMEs) and reviewers. While the machine will be able to generate a first draft that is 70 per cent accurate, it will never be able to provide the narrative or tell a story that only a human is capable of. In addition, to guard against implicit bias as well as hallucination, a higher degree of data fact check will be needed.
But there is no doubt that a human + machine combination can be a truly formidable partnership in this field. New skills will be needed that we had never contemplated. Prompt engineering is now a legitimate skill set that is being taught in colleges, just like data scientists did not exist 5 years ago!
JOHN WARD: The shift towards GenAI necessitates a transformation in the skill sets required on the shop floor. The focus moves towards skills in digital literacy, AI management, and data analysis. To support this transition, companies must invest in comprehensive training and development programs that not only focus on technical skills but also emphasise critical thinking and problem-solving in a technologically advanced environment. Learning new skills such as prompt engineering will be crucial for the workforce.
ANDREE BATES: In light of these changes, the emergence of new high-skilled roles will be more significant, with a focus on skills in managing AI systems. People need to be able to understand what is possible and how to manage AI. There will be more focus on human skills such as strategy, communication and others so having a mix of an understanding of how to manage AI and the softer human skills will provide an advantage.
What regulatory hurdles exist for Gen AI-driven solutions in pharma? How can companies navigate these challenges effectively?
SAMEER: This is a clear opportunity area, for sure. While the Joe Biden administration did release an executive order in late 2023 to establish new standards for AI safety and security, the reality is limited progress has been made around guardrails to drive safe, secure, and trustworthy development of AI. Currently, companies are self-regulating and most of the work has been done on internal-facing documents. In the absence of clear guidelines on copyrights, privacy, security, etc., each company is making its own internal guidelines based on its risk appetite and understanding of the space. This clearly is non-sustainable and needs to be corrected so that clear governance can be established before things get out of hand.
JOHN WARD: As GenAI becomes more integrated into pharmaceutical manufacturing, it's crucial to navigate the associated regulatory and ethical challenges. Ensuring data privacy, securing informed consent for data use, and maintaining data integrity are paramount. Companies must establish clear protocols and work closely with regulatory bodies to ensure their AI solutions comply with all legal and ethical standards.
ANDREE BATES: As AI systems assume increasingly autonomous roles in drug design and discovery, regulatory bodies will confront novel challenges in safeguarding the safety of algorithmically generated therapies. Conventional drug approval protocols hinge upon rigorous clinical trials and full transparency throughout the developmental journey. Nevertheless, the extensive utilisation of vast datasets manipulated in intricate, less intuitive manners by generative AI models may curtail the extent of traceability and comprehensibility. Regulators must thus devise fresh benchmarks to authenticate the safety and effectiveness of AI-derived molecules and research propositions intended for human application, leveraging the existing evidence. Moreover, it’s necessary to devise regulations governing AI systems themselves and ensure adequate oversight when models are updated or retrained over time.
What ethical considerations are associated with the use of GenAI in pharma, particularly regarding privacy, consent, and data security?
SAMEER: There are myriad dimensions of risks and ethical considerations for the responsible use of GenAI. It starts with copyrights as these LLMs have been developed on public sources of information that in turn have been generated from protected sources. Privacy concerns are justified if we are not able to protect against the release of personal information or data (even if inadvertently). The output data is dependent on the quality of the prompt that is the input. Inherent bias can creep in. The machine could “hallucinate” and cite references that do not exist putting a high burden on data fact-checking. These are just a few!
JOHN WARD: The ethical deployment of Generative AI (GenAI) in the pharmaceutical industry is critically important, particularly concerning the issues of privacy, consent, and data security. As we integrate these technologies at AskGXP, we are acutely aware of the responsibilities that come with handling sensitive data and the potential impacts on patients and consumers.
ANDREE BATES: As generative models advance, the establishment of scientifically and legally endorsed protocols for certifying AI-generated pharmaceuticals is anticipated to emerge as a paramount obstacle. Confronting this challenge is imperative to maximise patient access to the transformative potential offered by AI-enhanced drug discovery endeavours.
What advice would you give to pharmaceutical executives and decision-makers looking to leverage GenAI effectively in their organisations?
SAMEER: Cautious optimism, the theme that I started this interview with, is possibly the most practical way to approach the shifts in the landscape that executives and decision-makers are being seized with. Controlled experiments within the confines of their secure environments will allow them to form their own view of what is possible or not, and more importantly the best way to achieve what is possible. Keeping an open mind and preparing their teams for change will allow the scepticism and nervousness to be gradually replaced with an embrace of the possibilities that lie ahead of us!
JOHN WARD: To leverage GenAI effectively, executives should focus on building a robust digital infrastructure and fostering a culture of innovation and agility. This includes setting clear goals for AI integration, ensuring transparency in AI-driven decisions, and maintaining open channels of communication with all stakeholders. Additionally, executives should prioritise ongoing training and development to keep pace with technological advancements.
ANDREE BATES: With an eye on the future, pharmaceutical companies that prioritise enhancing their AI capabilities alongside implementing rigorous governance and oversight mechanisms are positioned to spearhead the industry into an era driven by artificial intelligence. This strategic approach promises to catalyse groundbreaking medical advancements for the global benefit of patients.
"Generative AI (GenAI) is transforming the pharmaceutical industry by enhancing data insights, content creation, and HCP engagement. It holds the potential to shorten drug development timelines and reshape workforce roles. Despite regulatory, privacy, and ethical challenges, cautious optimism and controlled experiments are essential for leveraging GenAI's full potential in pharma". Sameer
"The future of GenAI in pharmaceutical manufacturing promises significant advancements in productivity, efficiency, and innovation. By embracing these technologies, companies can not only streamline operations but also foster a more dynamic and innovative workplace. The key to successful integration lies in balancing technology with human insight and ensuring ethical use of AI". John Ward.
Author Bio
SAMEER heads the Enterprise Medical Solutions business at Indegene. He has 25 years of experience in the pharmaceutical industry and technology-led healthcare solutions provider sector. Prior to joining Indegene, Inc., he was associated with GlaxoSmithKline Pharmaceuticals Limited, Ranbaxy Laboratories Limited and Sudler & Hennessey Private Limited. He holds a bachelor of science degree and a bachelor of science (technology) degree, both from University of Mumbai and master of management studies degree from the Narsee Monjee Institute of Management Studies, Mumbai.
JOHN WARD serves as the Director of Ask GXP and ServBlock, with a robust background spanning over 14 years in pharmaceutical operations and advanced technology applications. At Ask GXP, John leads the adoption of Generative AI (GenAI) technologies, focusing on transforming processes across pharmaceutical manufacturing and significantly enhancing knowledge management systems. His leadership ensures the strategic integration of GenAI to streamline document generation, automate compliance tracking, and optimise data-driven decision-making processes.
In his role at ServBlock, John harnesses blockchain technology to safeguard data integrity and security within pharmaceutical supply chains, ensuring transparency and reliability.
DR. BATES Has spent over 20 years working specifically in pharma Artificial Intelligence. She brings blended expertise in Artificial Intelligence (AI), Pharmaceuticals, and Strategic Implementation of AI. Dr. Bates has led AI projects for numerous pharmaceutical in clinical trials market access, regulatory, medical affairs, and sales and marketing. (https://www.linkedin.com/in/dr-andree-bates).
