Event round-up
24 Sep 2025

Event Round-up | King’s Fund Industry Briefing | More than just hype: how emerging AI is assisting health & social care

Introduction 

AI in health and social care is shifting from pilot projects to practical, patient-facing use. In our recent techUK industry briefing with the King’s Fund, Pritesh Mistry walked through where value is landing today alongside the foundations needed for safe, scalable adoption. Below is a concise readout of the key insights, opportunities and constraints highlighted in the discussion. 

 
Watch the recording  

What’s happening 

The NHS 10-Year Plan signals an AI-powered NHS App and an ambition for “AI-enabled hospitals” by 2035. Realising that vision hinges on basics: working devices, robust connectivity, sensible cloud/on-prem mixes, accessible data of sufficient quality, and ongoing post-deployment monitoring to catch model drift. 

Where AI is helping now 

Quality & diagnostics 

  • Imaging & measurement: ML tools for segmentation and repeatable measurements (e.g., foetal ultrasound). 

  • Endoscopy support: Generative AI improves reporting and offers in-situ coaching to reduce missed findings. 

  • Unstructured EHR text: NLP flags patients needing intervention (e.g., repeated presentations, non-response to meds). 

Training & skills 

  • Drafting curricula and creating tailored learning materials. 

  • In-procedure coaching embedded in devices. 

  • Analysing recorded consultations to feed CPD with objective feedback. 

Workforce support 

  • Ambient AI scribes capture notes and codes, easing cognitive load and improving clinician–patient rapport. 

  • Drafting letters/assessments and summarising MDT meetings and FOIs to cut admin. 

Safety 

  • Automated call auditing in ambulance services (with caution on false positives). 

  • Peri-operative decision support for complex anaesthesia. 

  • Analytics aligned to PSIRF for incident learning. 

Access & experience 

  • Chat/IVR and website assistants for routine queries. 

  • Predicting DNAs to trigger targeted outreach or flexible appointment offers. 

  • Tailoring patient information by literacy/language and using utilisation data to plan pop-ups or relocate services. 

Operational productivity 

  • Automating repeatable measurements and documentation. 

  • Post-discharge automation: e.g., cataract follow-up calls with escalation to nurses (positive pilot feedback). 

What still needs fixing 

Digital foundations 

  • Device scarcity and locked-down features (cameras/mics). 

  • Network constraints for imaging-heavy workflows. 

  • Rising cloud storage costs driving hybrid architectures. 

  • Hard-to-access EHRs/document stores and variable data quality. 

  • IT teams stretched between uptime and innovation support. 

  • Need for robust post-deployment monitoring and drift detection. 

Governance & scale 

  • Growth of provider clusters and innovation boards to triage and share adoption. 

  • Market competition encourages innovation but creates some duplication. 

  • Regulators (e.g., MHRA) are adding capacity and sandboxes, but throughput vs demand remains an open question. 

Equity & beyond acute 

  • Community and mental health see strong fit for ambient voice, scheduling and outreach, but often lack IG/clinical safety resourcing and change capacity. 

Trust & accessibility 

  • AI can improve accessibility (multi-format content), yet public trust is uneven. Local validation, transparency and user testing are essential. 

 

Return to listing