Clinicians are holding back from using AI tools because no one can tell them who will be held responsible if the technology causes patient harm. At the HLTH Europe conference in Amsterdam on June 16, delegates heard that the unresolved question of liability - whether fault falls on the hospital, the clinician, or the software developer - remains one of the largest obstacles to adoption in clinical settings.
The legal vacuum is not theoretical. AI-based diagnostic support, treatment planning algorithms, and administrative triage systems all carry the possibility of error. When those errors lead to a missed diagnosis or an inappropriate treatment, the chain of accountability often fractures. Hospitals argue they rely on regulatory clearance. Clinicians point to the algorithm's recommendation. Developers claim the tool was meant to assist, not replace, human judgment.
Why uncertainty stalls adoption
This ambiguity has a chilling effect. Without clear rules, healthcare organizations hesitate to deploy tools that could improve efficiency and outcomes. Clinicians, already cautious about professional liability, are reluctant to incorporate AI into workflows when they cannot predict how a court or licensing board would assess their actions. The result is a slower path from regulatory approval to bedside use.
For healthcare professionals, understanding the technology itself is a first step toward managing that risk. Resources such as AI for Healthcare courses can help build the foundational knowledge needed to evaluate when and how to trust an algorithmic output.
The European regulatory context
Europe's AI Act and the Medical Device Regulation set boundaries for high-risk AI systems, but they do not yet provide a clear liability framework for individual patient cases. The panel discussion highlighted that while regulators are moving to classify software, courts have little precedent to guide them. This leaves each stakeholder group trying to shift exposure through contracts, disclaimers, and insurance policies - a patchwork that does not inspire confidence at the clinical level.
Why this matters for healthcare professionals
The liability gap means that clinicians who use AI today are operating in a gray zone. Until legislation or case law provides clarity, the safest approach is to document every decision point where AI played a role, understand the tool's limitations, and advocate for institutional policies that define shared responsibility. The technology will keep advancing; the legal scaffolding needs to catch up before adoption can match the ambition.
Your membership also unlocks:
