AI is moving beyond chatbots. A new generation of AI agents can take actions on their own: calling on tools, connecting to other systems and making decisions to complete a task. In healthcare, that could mean an agent that reviews patient safety incidents, summarises them and emails the right people, without a human checking each step.
The potential is real, and so are the risks. When an agent can act on its own, it isn’t always clear what it will do, what it can access, or who is accountable when something goes wrong. For healthcare organisations, those questions sit at the heart of patient safety.
Newton’s Tree and Curistica are bringing together clinical safety experts and NHS digital leaders for a live discussion on what agentic AI means for healthcare and how to deploy it safely.
What we’ll cover
- Responsibilities: what healthcare organisations are responsible for when they deploy agentic AI, and where those responsibilities lie between the organisation and the supplier.
- Governance: what goes wrong when governance is weak, and why it matters more for agents than for traditional AI.
- Hazards and controls: real examples of how agentic AI can fail in practice, and the controls that reduce the risk.
- Live Q&A: put your questions to the panel.
Meet the panel
- Duncan Dewhurst, Chief Digital and Transformation Officer, Buckinghamshire NHS Trust
- Keith Grimes, Founder and CEO, Curistica
- Youssof Oskrochi, Head of Safety & Data Protection, Curistica
- Haris Shuaib (Chair), Clinical Safety Officer and CEO, Newton’s Tree
Who should attend
This session is for healthcare leaders, clinical safety officers, digital and IT teams, and anyone thinking about bringing agentic AI into their organisation.
Want to continue the conversation in person? Join us on Thursday 26 November in central London for an evening of debate on whether agentic AI belongs in the NHS, followed by drinks and networking. Places are limited, sign up here
