Innovation in the NHS: Personalised Medicine and AI Inquiry Launched
The House of Lords Science and Technology Committee has opened an inquiry into innovation in the NHS, using personalised medicine and AI as test cases for why translation often stalls-and how to fix it. Written evidence is invited by 23:59 on Monday 20 April 2026.
This matters to anyone in UK life sciences turning research into patient impact. Genomics, AI-driven analytics, and cell and gene therapies like CAR T are moving from research programmes into real services, but deployment at scale is still uneven and costly.
What the inquiry will examine
- The current state of science behind personalised medicine, the biggest gaps, and how AI contributes across prevention, diagnosis, and treatment.
- The research infrastructure needed in the UK, from secure data environments and compute to trials capacity and biomanufacturing.
- How well the UK translates scientific strengths into validated tools-and what will keep the country competitive.
- How proven innovations can be deployed across the NHS, and the systemic barriers that slow or block adoption.
- The suitability of regulatory frameworks and where they should be improved.
- Whether appraisal and commissioning models fit personalised medicine and algorithmic tools.
- How Government can strengthen feedback loops between medical research, industry, and the NHS.
A full list of questions is available in the Call for Evidence.
Chair's perspective
Lord Mair CBE, Chair of the Committee, said:
"Advances in AI and genomics are creating the prospect of truly personalised medicine across prevention, diagnosis and treatment. Our inquiry will use personalised medicine as a case study to explore a broader question: why does the NHS struggle to adopt the UK's life sciences innovations, and what could be done to fix that?
"As the NHS plans to use new developments in genomics, AI, and personalised medicine, our inquiry will seek to establish the state of the science and technology in this area and understand where patients might benefit from near-term developments.
"We will examine the gap between early-stage research, clinical trials and NHS-wide delivery, looking at blockages in the system slowing progress, including procurement processes, clinical pathways, and the role of regulators and professional bodies.
"We will also examine how the fragmentation of the overall NHS structure contributes to the uneven deployment of innovation, how the costs of personalised treatments can be reduced, and the clinical academics and clinical trials infrastructure needed to rapidly deploy innovations within the NHS."
Why this is timely for researchers
Genomic medicine is accelerating, from variant interpretation to therapy development. AI models promise faster analysis and decision support, but evidence standards, data access, and NHS integration remain sticking points.
For context on current NHS genomics delivery, see the NHS Genomic Medicine Service. For population-scale research assets that enable discovery and validation, review UK Biobank.
What high-quality submissions should cover
- State of the science: Clear view of clinical areas closest to impact (e.g., oncology, rare disease, pharmacogenomics), critical unknowns, and realistic near-term milestones.
- Data and compute: Access to linked, longitudinal datasets; use of trusted research environments; interoperability with FHIR; reproducible pipelines; compute requirements and costs.
- Validation and safety: Prospective studies, external validation, bias assessment across demographics, MLOps and model monitoring in clinical settings, failure modes, and update governance.
- Clinical pathways: Where a tool or therapy fits, who acts on outputs, workflow changes, and time-to-result. Include service design and training needs for clinicians and analysts.
- Regulation and standards: Practical suggestions for proportionate, risk-based oversight across diagnostics, software as a medical device, and ATMPs, including post-market surveillance.
- Appraisal and commissioning: Fit-for-purpose methods for personalised therapies and AI tools-beyond average-effect models-covering subgroup effects, companion diagnostics, and outcomes-based payment.
- Manufacturing and logistics: For cell and gene therapies, address QC, release testing, chain of identity/custody, and options to lower per-patient cost (platform processes, shared infrastructure).
- Equity and access: Safeguards to avoid widening disparities; inclusion in datasets and trials; regional deployment plans that match capability and workforce.
- Procurement and incentives: How to buy AI and personalised interventions (modular contracts, pay-for-performance, sandboxing), and how to align incentives across ICSs and providers.
- RWE and feedback loops: Registries, outcome tracking, and data flows that send real-world performance back to researchers, regulators, and commissioners.
- Security and privacy: Practical approaches for information governance, privacy-preserving analytics, and patient trust.
Practical tips for contributors
- Be specific: reference datasets, trial identifiers, and quantified outcomes where possible.
- Propose pilots that could run in 6-12 months with clear endpoints, budgets, and success criteria.
- Show total system value: clinical outcomes, capacity released, avoided harms, and costs over the full pathway.
- Address scalability from the outset: workforce, training, integration with EHRs, and support models.
- Include a plan for continuous evaluation and safe iteration after deployment.
Who should respond
Academic and industry researchers, clinical leaders, data scientists, bioinformaticians, health economists, ATMP manufacturers, regulators, commissioners, and patient groups with direct experience of AI or personalised medicine development and deployment.
Timeline
Deadline for written evidence: 23:59 on Monday 20 April 2026.
Resources
- AI for Healthcare - practical training and tools for applying AI to clinical data, model validation, and deployment.
Your membership also unlocks:
