Who's Liable When AI Gets It Wrong? | Robert Kellar KC on AI, Law and Liability in Regulated Professions

When an AI tool contributes to a clinical decision and a patient is harmed, who is liable? The clinician who acted on it, the organisation that deployed it, or the company that built it? Right now, English law cannot give a clean answer, and every organisation adopting AI is operating inside that uncertainty. In this episode of The TechDental Podcast, Dr Randeep Singh Gill sits down with Robert Kellar KC, a barrister at 1 Crown Office Row specialising in clinical negligence, professional discipline and regulation, public law and personal injury. Robert has spent recent years litigating high-value clinical and dental negligence claims, and was writing and speaking about AI and legal liability for medical practitioners before most people had heard of ChatGPT. This is a high-level conversation about legal governance and the reality of AI liability in regulated professions. The case study is dentistry. The framework applies to medicine, law, finance, and any high-stakes field where a regulated human is accountable for a decision an algorithm increasingly shapes. Dr Randeep Singh Gill and Robert Kellar KC discuss why the law looks to the human practitioner first and the technology company third, how AI could dilute or displace the 70-year-old Bolam standard of negligence, the two traps every clinician now sits between, automation bias on one side and the risk of wrongly overriding the AI on the other, and why documented independent judgement is the single best legal defence. They also cover the questions every leader should ask an AI vendor before signing, why a 1987 product liability law may or may not treat AI software as a product, why regulation gives innovators certainty rather than stifling them, how to use the EU AI Act as a procurement gold standard, and the governance every organisation should build before deployment. Essential listening for dentists, doctors, DSO operators, healthtech founders, compliance leaders and investors navigating AI adoption, clinical governance and professional liability. **Please note:** This episode is for general information and discussion only. It is not legal advice. These issues are highly fact-specific, so take qualified legal advice before acting, and do not rely on AI as a substitute. **In this episode:** Why the liability chain runs to the practitioner first, then the organisation, then the developer How AI could dilute or reconceptualise the Bolam standard of clinical negligence Automation bias and the risk of disagreeing with AI: the vice closing from both sides Why documented judgement is a clinician's primary legal defence The two questions to ask any AI vendor before deployment Whether AI software counts as a "product" under the Consumer Protection Act 1987 Why regulation provides certainty for innovators rather than blocking them Using EU AI Act compliance as a procurement and governance benchmark The governance framework to build before AI reaches a patient Who Robert predicts will stand in the dock when the first case reaches court **About the guest:** Robert Kellar KC is a barrister at 1 Crown Office Row specialising in clinical negligence, professional discipline and regulation, public law and personal injury. He is a board advisor and a writer, speaker and thinker on medicine, AI and web3. LinkedIn: https://www.linkedin.com/in/robertkellarbarrister1crownofficerow/ Chambers: https://www.1cor.com/london/ **Read the full written analysis:** https://www.techdental.com/insights