How GenAI will transform life sciences in 2025
The life sciences industry is in a position to help humanity, perhaps more than ever, with the introduction of generative AI (GenAI). Lengthy drug development timelines and complex regulatory hurdles are two of the most longstanding challenges in this space, and this new technology can significantly support and accelerate these processes for biotech and pharmaceutical companies.
In 2024, numerous pilot programmes transitioned from experimental stages to production, with Google Cloud customers accelerating drug discovery timelines, streamlining clinical trial processes, and establishing the building blocks to delivering more personalised patient care. In fact, we recently conducted a study which found 62% of healthcare and life sciences execs have already moved GenAI use cases into production, with 74% seeing ROI from their GenAI investments on at least one use case.
Now, in 2025, we expect the life sciences industry will scale its use of GenAI with greater purpose. As these applications continue to deliver ROI, GenAI adoption across the industry is set to increase, unlocking new potential at the intersection of science and technology.
We’re seeing four trends drive this adoption, including multimodal AI, AI agents, intuitive search, and AI-powered consumer experiences. In the following article, I’ll explore each of these trends and why they’re important to biotech and pharmaceutical companies in 2025.
1. Multimodal AI: Unlocking deeper insights and new innovations
The success of AI models – particularly in the life sciences industry – depends heavily on the richness and diversity of underlying data. These models must be capable of processing vast, complex data sets, including a mix of images, texts, charts, genomic information, and medical records. To deliver on this promise, the underlying AI models must be multimodal – meaning they can integrate and analyse data from multiple formats. By providing a more comprehensive and nuanced understanding of scientific data, multimodal AI can not only accelerate drug discovery, but also enhance diagnostics.
For example, Bayer is harnessing AI to revolutionise drug discovery by analysing extensive datasets and automating critical tasks, accelerating the journey toward new medicines. Leveraging generative AI, Bayer built solutions that can automate the completion of up to 80% of regulatory dossiers, streamline regulatory workflows, and speed up access to medications for patients. In oncology, Bayer is using synthetic images generated from histology data to overcome the challenges of limited training data, especially in rare diseases. These synthetic images have shown great promise in enhancing diagnostic capabilities in oncology and radiology, helping fill crucial data gaps in these fields.
As an industry, we want to be able to identify the right treatment for the right patient at the right time. We no longer accept that it often takes 10 to 20 years for personalised medicine solutions to come to market. Multimodal AI will help reveal patterns and predictions that were previously beyond reach, driving faster diagnoses, more efficient drug development, and deliver a new era of personalised medicine.
2. AI agents: Automating tasks and optimising workflows
In 2025, AI agents will play an increasingly transformative role across the life sciences industry. These advanced AI tools are capable of automating complex tasks, from genomic data analysis and clinical trial design to report generation and support for commercial activities. We’ll see AI agents deployed widely to streamline workflows, enhance decision-making, and boost overall productivity for employees.
AI agents may be used to optimise clinical trials by swiftly analysing patient data, pinpointing ideal recruitment locations, and automating the assessment of trial outcomes. This level of automation holds the potential to significantly reduce the time and costs associated with clinical trials, ultimately accelerating the process of bringing new medicines to market. As these agents take on repetitive, data-intensive tasks, researchers and clinicians will be able to focus more on strategic work, thus pushing the boundaries of innovation in patient care and drug discovery.
For example, Exscientia leverages Google Cloud GenAI capabilities to accelerate drug discovery through a Design-Make-Test-Learn (DMTL) cycle. These AI agents analyse datasets to identify drug targets and design novel molecules using GenAI and active learning algorithms. This collaboration enables faster drug discovery, improves drug efficacy and safety, and fosters innovation by exploring new chemical spaces and identifying previously unknown drug targets.
3. Intuitive search: Transforming research and facilitating new discoveries
Intuitive search, driven by natural language processing (NLP), is set to transform how researchers and healthcare professionals access information in 2025. Currently, conducting clinical trial research or literature reviews often involves manual, time-consuming searches. By understanding the nuances of both human language and complex scientific terminology, intuitive search engines can deliver more accurate, relevant results – even from incomplete queries or those with medical abbreviations and specialised terms.
Companies like Ginkgo Bioworks are using generative AI enterprise search to tap into their extensive codebase, which includes both labelled and unlabelled data. With search technology, Ginkgo can efficiently pinpoint relevant data from past experiments and academic research. This approach enhances their ability to access critical information quickly, which is invaluable for launching new programmes and driving research forward.
With intuitive search, researchers will be able to use natural language to locate relevant studies, identify similar compounds, and stay updated on the latest advancements in their fields. This enhanced capability will not only accelerate research workflows, but also facilitate new discoveries by making critical information accessible more quickly and efficiently. Ultimately, this technology will enable life sciences professionals to harness vast knowledge repositories, supporting faster innovation and more informed decision-making across the industry.
4. Customer experience: Enhancing patient engagement
Today, an increasing number of patients take an active role in managing their healthcare decisions. Consumers expect life sciences companies to equip them with accessible tools and resources to help them make informed choices. GenAI-powered chatbots, and virtual assistants can engage with patients directly, providing educational information about conditions, treatment options, and how to find support services.
Additionally, GenAI can help life sciences companies tailor communications to specific patient needs and preferences, creating a more personalised and engaging experience. By addressing individual concerns and offering relevant guidance, AI-driven customer experiences can foster greater trust and satisfaction, helping companies meet evolving patient expectations.
Looking ahead: Increasing protections and establishing trust
Over the next three to five years, we’re likely to see increasingly sophisticated applications of AI across the life sciences, which have the potential to drive faster drug development, deliver more personalised treatments, and ultimately improve patient outcomes.
I believe the technology to power a fully AI-driven drug discovery process may already exist today. However, realising its full potential will require the industry to continue its strong focus on ethical considerations, data privacy, and cross-industry collaboration.
To ensure that AI advancements benefit all stakeholders, it is essential to prioritise building trust across the industry through education, transparency, and close partnerships with regulatory bodies. By working together, we can create a future where personalised medicines and proactive healthcare become the new standard.
