The Future of AI in Healthcare: Jackie Hunter on Innovation, Ethics and Adoption
Jackie Hunter is one of the UK's most trusted voices on pharmaceutical innovation and ethical AI. She has led teams at GlaxoSmithKline, the Biotechnology & Biological Sciences Research Council, and BenevolentAI, and now chairs Brainomix Ltd, the Stevenage Bioscience Catalyst, and Biocortex Ltd. Her work has been recognised with a CBE and a place on Forbes' list of Top 20 Women Globally Advancing AI Research. She is also a Fellow of the Academy of Medical Sciences and a Visiting Professor at Imperial College London.
As a keynote speaker with The AI Speakers Agency, Jackie shares how AI can improve drug discovery, clinical services, and health systems-without losing sight of ethics, transparency, and patient impact.
Where AI already delivers value
"Successful artificial intelligence is already shaping care," Jackie notes-specifically in radiology, pathology, patient triage, and remote home care. These aren't pilots on the fringe; they're live services improving throughput and consistency. The real hurdle now is adoption at scale, not proof of concept.
That means commitment across a full pathway or service line, clear senior sponsorship, and-most importantly-frontline engagement. "AI in its implementation is not just technology, it is also a social science." If clinicians and operators are part of the build, the workflows stick.
Implementation: start with the problem, not the tool
Jackie's first question to any team: what problem are you solving? Are you making an existing process faster and more reliable, or are you redesigning the process end-to-end? Incremental upgrades are easier to ship; disruptive changes can deliver bigger gains but require deeper buy-in and change management.
She points to examples in pharma: some organisations integrate AI across the value chain, while others keep it to isolated pilots. The pattern is clear-integrated beats isolated.
Make AI part of the craft, not a side project
Jackie draws a parallel to the 1990s: molecular biology started as a specialist department and became embedded in every team. AI should follow the same path. Teams that "put AI in a box" separate from domain expertise tend to see shallow results.
The move is to embed data scientists with clinicians, pharmacists, and operational leaders. Shared ownership beats handoffs.
Ethics that hold up in clinic
- Data quality and representation: Training sets must reflect real patient populations across ethnicity, socio-economic status, and comorbidities. Be extra careful when augmenting with synthetic data.
- Bias beyond data: Watch for bias in interpretation and downstream use, not just in the inputs. Governance needs to follow the full pathway.
- Transparency: Clinicians need to understand how recommendations are produced. Supervised models are easier to explain; for unsupervised methods, dig into how outputs are derived and documented. This aligns with ongoing work on explainability highlighted by leaders such as Demis Hassabis at Google DeepMind. Learn more
- Standards and accountability: Use established frameworks for safe deployment and monitoring. The UK's guidance on data-driven health tech is a useful reference point. See the code of conduct
Borrow what works from other industries
Healthcare often moves carefully, and for good reason. Still, other sectors show that rapid adoption is possible with the right incentives, education, and communication. Workforce enablement is the lever-people need to see how AI frees time for complex cases and deeper patient engagement, rather than replacing roles.
Think agile practices, fast feedback loops, and visible wins. Keep the ethics and regulatory standards intact while clearing the path for measured progress.
Open innovation: build, partner, or spin out?
Open innovation sits at the core of Jackie's work, including the early vision for the Stevenage Bioscience Catalyst. The principle: decide what is truly core to build, and where partnering or acquiring is faster and smarter. Don't reinvent a platform if a proven solution exists.
Also, don't let valuable IP collect dust. If it's not strategic inside your organisation, spin it out and create value with the right business model.
Action steps for healthcare leaders
- Pick 1-3 use cases with clear ROI (e.g., image triage, coding automation, readmission risk).
- Embed cross-functional teams: clinical, operations, data science, IT, and legal at the same table.
- Run a data audit: coverage, quality, bias checks, lineage, and consent.
- Start with explainable approaches; document decision logic and limits from day one.
- Plan the workflow: who acts on the output, how, and what safety nets exist.
- Measure what matters: accuracy, time saved, equity impact, patient outcomes, and clinician satisfaction.
- Invest in training and change management; appoint clinical champions and superusers.
- Adopt continuous monitoring and model refresh cycles; treat models like living systems.
- Review procurement and IP strategy for build/partner decisions and potential spinouts.
About Jackie Hunter
Jackie Hunter, CBE, is a leader in pharmaceutical R&D and ethical AI. She chairs the boards of Brainomix Ltd, the Stevenage Bioscience Catalyst, and Biocortex Ltd, and previously served in senior roles at GlaxoSmithKline, the BBSRC, and BenevolentAI. She is a Fellow of the Academy of Medical Sciences, a Visiting Professor at Imperial College London, and was named by Forbes as one of the Top 20 Women Globally Advancing AI Research.
She is a keynote speaker with The AI Speakers Agency, frequently addressing how to adopt AI responsibly across drug discovery, clinical services, and life sciences operations.
Keep your teams current
If you're building capability across clinical, data, and operations teams, curated AI learning paths can help. Explore role-based options here: Complete AI Training - Courses by Job.
Your membership also unlocks:
