A new study published in Animals has set out the priority areas where artificial intelligence, or AI, could be used to improve the health and welfare of horses, ponies, donkeys and their hybrids in the UK. The paper brings together expert opinion on where AI could offer the greatest benefit, what the main risks are, and what needs to happen if these tools are to be used responsibly and effectively.
The research, led by Philippa L. Young and colleagues at the University of Nottingham and World Horse Welfare, used a workshop and Delphi survey process to gather views from stakeholders across the equid sector. In total, 41 people attended the initial workshop, and the wider process produced 134 statements, with consensus reached on 106 of them.
Among the clearest messages from the study was that AI development should be guided by real welfare need, not just technological possibility. Participants identified unmet ethological needs, poor management practices, and gaps in owner or carer knowledge as some of the most important welfare concerns affecting UK equids. Priority areas for AI development included tools to assess equid wellbeing, monitor individual animals, track equid populations, and support earlier recognition of health and welfare problems.
But the paper also makes clear that AI is not a simple fix. Stakeholders raised concerns around poor-quality or biased data, difficulties validating AI models, lack of public understanding, ethical issues, and uncertainty over how AI outputs should be interpreted. There were also concerns that over-reliance on AI could lead to over-treatment or a loss of practical evaluation skills within the equine sector.
Importantly, the study does not argue for more AI at any cost. Instead, it calls for evidence-based, unbiased systems, better education for users, stronger collaboration, and clearer approval or regulatory processes. The authors say these findings should help guide future funding and development, so that AI tools are targeted towards the areas where they are most likely to make a meaningful difference to equid welfare.
The authors describe this as the first study to identify stakeholder consensus on where AI is likely to have the greatest benefit for equid health and welfare, as well as the barriers and practical solutions that need to be addressed along the way. At a time when AI use is expanding rapidly across animal health and welfare, the paper offers a timely framework for thinking about how these technologies should, and should not, be used in the equine world.
