Exploring how causal AI can support safer, more explainable and more effective decision-making across healthcare and social care.

This event will bring together researchers, healthcare and social care professionals, NHS and public-sector stakeholders, technology developers, policymakers, and people with lived experience to explore one of the most important emerging opportunities in responsible AI: using causal AI to support better decision-making across healthcare and social care. Current AI systems are increasingly effective at prediction, information retrieval, summarisation, and pattern recognition. However, many important care decisions require more than identifying correlations. Professionals need to understand why an outcome occurred, what factors contributed to it, what evidence may be missing, and whether a different intervention could have changed the result. These questions are particularly important in areas such as falls, medication-related incidents, deterioration, safeguarding, care coordination, and clinical risk assessment. Causal AI offers a promising route towards addressing this gap. By combining causal reasoning with responsible and explainable AI, future systems could help professionals move from simply asking “What is likely to happen?” towards questions such as “Why did this happen?”, “What caused this outcome?”, and “What might have prevented it?” The workshop will also examine the growing role of agentic AI in healthcare and social care. These systems can support multi-step tasks such as collecting information, retrieving relevant records, identifying missing evidence, summarising incidents, and supporting organisational learning. However, their use in sensitive care environments introduces important challenges around uncertainty, human oversight, explainability, fairness, accountability, privacy, robustness, and safe escalation. For healthcare and social care organisations, this creates a significant opportunity. Causal AI could support more meaningful clinical and care decision support, improve incident investigation, strengthen organisational learning, and help professionals reason about interventions rather than relying only on predictions. At the same time, the technology must be developed with strong safeguards so that AI supports rather than replaces professional judgement. This workshop will explore where causal AI can provide measurable value, what evidence and data are required, what barriers may slow adoption, and how academia, healthcare, social care, industry, and public-sector organisations can work together to develop trustworthy and practically useful solutions.

Funded by CHAI HUB. (https://www.chai.ac.uk/)

Description supplied by the event organiser.