A short-term AI strategy for 2025 – 2026 from Midlands Partnership University NHS Foundation Trust (MPFT) has been published, as the trust looks to learn from the first 12 months of AI deployment ahead of a full AI strategy update. It prioritises “high-impact, lower-risk use cases, with rigorous evaluation, human-in-the-loop safeguards and strong governance” across ambient AI, generative AI, agentic AI, and automation.
The trust’s AI work will be informed by seven AI principles covering safety, confidentiality, research, evaluation, partnership, ethics, transparency, and data quality. All adopted technologies will be assessed for compliance; every AI initiative will be evidence-led and risk-aware; and working with educational institutions and “trusted suppliers” will supplement MPFT’s AI expertise. “We will engage with people impacted by AI solutions and ensure effective co-design to enhance health and care experiences, address needs, and build trust in AI technology,” the trust states. “For all AI initiatives we will co-produce with our staff, service users and carers, to ensure the technology works for people within the communities we serve.”
Highlighted are some of the AI technologies already making an impact at MPFT, with Co-Pilot and Intelligent Recap reportedly reducing staff workload by summarising non-clinical meeting content and identifying key actions; the Rio Virtual Assistant supporting service users to view and manage appointments, optimising appointment slots; and the Psyomics platform improving triage by providing pre-assessment in adult mental health.
Also mentioned are the trust’s Limbic chatbot which supports Shropshire, Telford & Wrekin Talking Therapies by triaging self-referrals. Rio Smart Notes uses ambient AI to summarise consultations, T-Pro voice dictation’s AI-powered voice-t0-text functionality offers dictation with intelligent summarisation, and robotic process automation is used for “repetitive tasks” such as death registrations and recruitment activities.
Over the next year, the focus at MPFT will be on using ambient AI to improve consistency in note-taking and reduce the burden on staff, using capabilities such as voice-to-text transcription. Generative AI and large language models will be used to support staff in non-clinical activities like drafting documentation, and RPA will help with admin processes “with a focus on referral processes and workflow improvement”. AI agents will also handle appointment booking queries and coordinate activity automatically based on service user requests.
The trust’s wider approach is “fail fast”, it shares. “This mindset allows us to explore the potential of AI in a controlled, low-risk way, starting small, iterating fast, and scaling only what delivers real value as demonstrated through benefits evaluation. By learning rapidly, we will be making smarter decisions that benefit our service users, staff, and services.”
To ensure rigour in its use of AI, MPFT is co-developing and evaluating AI tools with Keele University, whose technical expertise will help support the trust in validating performance, monitoring bias, and ensuring AI aligns with realities of frontline care. Efficiency and sustainability considerations cover only using AI where it can make the biggest impact on improving outcomes or enabling new capabilities, and using it to help reduce unnecessary appointments, travel, and duplication to lower emissions.
MPFT commits to building AI literacy across its workforce using training, e-learning, workshops, and clear communications, with a framework to be designed to offer role-specific training “so staff and service users feel confident and empowered to use AI appropriately”. An AI champions network will be established, it shares, and every AI pilot will be “treated as a workforce and service user learning opportunity, gathering insights to adapt training and deployment”.
As MPFT moves into the pilot and evaluation phase from October 2025 to March 2026, it will be monitoring initial AI pilots, collecting baseline and post-pilot data, refining training based on real use, researching AI decision support potential, and looking at a Call Handling and Admin Automation Transformation Programme (CHAAT-P). The scale and embed phase from April to September 2026 will move on to integrate successful pilots into BAU workflows, launching an AI dashboard for live tracking, and preparing the next wave of use cases across AI diagnostics, AI decision support, and AI triage.
Later in 2026, learnings will be used to plan and prioritise for AI in 2027 and beyond, with a longer term plan expected to be published in September 2026.
Wider trend: AI and AI strategy from across the NHS
In April, NHS England issued guidance on the use of AI-enabled ambient scribing products in health and care settings, to support chief information officers and chief clinical information officers when introducing this technology. The guidance aims to support business case development, risk assessments, governance, data protection impact assessments and evaluation and monitoring. A key part of the framework covers a series of considerations, posing questions to clarify the product functionality, outputs for transcription, outputs for downstream tasks and data and system considerations.
Alder Hey Children’s NHS Foundation Trust published their own AI Strategy earlier this year, outlining current and future AI work along with plans for benefits realisation, implementation, AI workforce development, infrastructure and data architecture. Driving the transformation are four key themes: enhancing children and young people centred care; empowering colleagues and freeing up time using intelligent automation, AI assistants, and smarter workflows; transforming outcomes for children and young people by delivering precision care through AI-optimised pathways, predictive analytics, and remote monitoring tools; and revolutionising paediatric diagnostics with “cutting edge” innovation.
Great Ormond Street Hospital for Children NHS Foundation Trust (GOSH) has also released its AI strategy, highlighting the “unprecedented opportunity” AI presents for revolutionising healthcare delivery and enhancing clinical outcomes, driving efficiencies, reducing clinician burden, and improving patient care. In year one, governance, reporting, and monitoring structures will be developed, and a gap analysis will be undertaken to understand current capabilities, infrastructure, and use cases. A roadmap will be created based on the HIMSS Analytics Maturity Assessment Model (AMAM) framework, and the formation of an analytics council and a Department of Advanced Data Analytics and AI will support governance work. The trust also plans to launch pilots of AI solutions across diagnostics and administration, and to roll out AI literacy training for staff.
