Last month, CHIA’s Trustworthy Artificial Intelligence Lab (TRACE), led by Assistant Professor Umang Bhatt, joined health care and AI leaders from across the UK to launch TrustX, a partnership designed to verify, deploy, and test AI agents safely across health and care settings. This partnership is the first of its kind focused on scientific, auditable, and scalable deployment of Agentic AI within the NHS and social care settings. TrustX supports NHS’s “Fit for the Future” 10 Year Health Plan for England, which calls for a shift towards the widespread use of AI by creating methods for safely evaluating Agentic AI in health and care settings.
TrustX officially launched at University of Cambridge’s Trinity Hall on December 9, bringing together academics, government leaders, NHS England, and members of life sciences and AI companies. Professor Bhatt said of the TrustX partnership, “Deploying AI in health and care should be earned through evidence. TrustX allows us to bring scientific rigour to the evaluation of autonomous AI agents, linking human-centered research with real clinical and operational environments. This is key to building AI agents that behave reliably, align with human values, and can be deployed at scale with confidence.” TRACE PhD and MPhil students Drew Calcagno and Shin Zert Phua were crucially involved in the planning of the event and in the TraceX partnership.
TrustX aims to deliver the following:
- A front door for AI agent deployment across the NHS and social care, including:
-
-
- Scoring and verification of existing Agentic AI systems
- Skunkworks evaluation to create a collaborative space to determine which NHS and social care problems are appropriate for Agentic AI
- Support to build new AI agents or improve them to mitigate risks
- Real-world evaluation against productivity and cost-effectiveness metrics
-
- An open-source Agentic AI Trust Score to accelerate transparency and adoption
- A national collaboration environment to assemble leaders from the NHS, academia, industry, civil service, and research institutes
- Partnerships with NHS provider and social care sites to test live Agentic AI deployments, beginning with early work already underway at Sussex Partnership Foundation Trust
- A sustainable and flexible funding structure, combining different funding approaches to support organisations, from startups to large suppliers, and the NHS and social care
CHIA’s Trustworthy AI Lab joins the following partners in this important work:
- NHS England South East
- Health Innovation Kent Surrey Sussex
- The Responsible AI Institute
- The King’s Fund
To read more about the TrustX partnership, please click here.