This exclusive interview with Jackie Hunter was conducted by Tabish Ali of the Motivational Speakers Agency.
Artificial intelligence is already being used across radiology, pathology, patient triage and remote care. Its wider impact will depend on how successfully healthcare organisations integrate the technology, engage clinical users and address concerns around bias, data quality and transparency.
Professor Jackie Hunter CBE is an innovation expert with an extensive career in pharmaceutical research, biotechnology and AI-led drug discovery. She is Chair of Brainomix, Biocortex and Stevenage Bioscience Catalyst, and previously served as CEO and Board Director of BenevolentAI and Senior Vice President at GSK.
In this exclusive interview with the AI Speakers Agency, Jackie discusses where AI is already delivering value in healthcare, how pharmaceutical companies can move beyond isolated pilots and the ethical questions that must be addressed as adoption grows.
Question 1: Where is AI already delivering value in healthcare, and what will determine whether it achieves wider clinical adoption?
Jackie Hunter: “Artificial intelligence is already shaping healthcare. It is being employed in radiology, pathology, patient triage and remote home care, with wider implications for health services.
“Realising its full potential requires a commitment to adoption across the healthcare landscape. Senior management buy-in is important, but users must also be fully engaged. Implementing artificial intelligence involves social science as well as technology.”
Question 2: How should pharmaceutical and healthcare businesses approach AI implementation, and what separates meaningful integration from isolated experimentation?
Jackie Hunter: “Businesses must first identify the problem they are trying to solve. They need to decide whether they want to make an existing process more efficient or transform it completely.
“For example, chemical synthesis and screening can be combined robotically to improve iteration and throughput. Radically transforming a process can be exceptionally valuable, but it is also much harder to implement.
“Implementation in the pharmaceutical industry is currently fragmented. Companies such as Amgen are integrating AI across the entire value chain. Others are conducting pilots and putting their toe in the water.
“The development of molecular biology provides a useful comparison. In the 1990s, we understood gene cloning and DNA, while RNA was becoming increasingly important in healthcare and pharmaceuticals. Companies had separate molecular biology departments. Molecular biology is now seamlessly integrated into everything, from patient stratification to target identification.
“To embrace the power of AI, companies must embed it within their domain expertise. Businesses that place AI in a separate box are less likely to succeed than those integrating it across the whole organisation.”
Question 3: Which ethical risks must organisations address when deploying AI in healthcare, particularly around data quality, bias and explainability?
Jackie Hunter: “The quality of the data is critical. When developing clinical trial algorithms, organisations must include a broad range of patient populations, ethnicities and socioeconomic groups to ensure the data is representative. This is especially important when synthetic data is used to enlarge training sets.
“The second ethical concern is bias. This includes bias within the data, its interpretation and the downstream application of AI.
“Transparency is also essential. Organisations must understand how AI models reach their conclusions.
“With supervised learning, this is relatively straightforward because you know how the algorithm has been trained. With unsupervised learning, the parameters behind its decisions may be less clear. Organisations must be able to examine and explain how an algorithm produces its recommendations.
“Demis Hassabis of Isomorphic Labs has said that one of his organisation’s aims is to develop AI systems that can explain how they reached their recommendations.”
