Liability ‘gap’ in AI health guidance – MPS
An organisation that provides support to 16,000 healthcare professionals in Ireland has welcomed draft guidance on the use of AI in healthcare settings but has expressed concern about the absence of any reference to liability.
Medical Protection Society (MPS) was responding to a public consultation on HIQA’s draft National Guidance for the Responsible and Safe use of Artificial Intelligence in Health and Social Care Services.
The organisation pointed out that AI providers sometimes included contractual clauses limiting or excluding their responsibility for the performance of their systems, adding that this was a significant issue in practice.
“The guidance should either address the issue directly or advise clinicians and services to avoid entering into contracts that absolve AI providers of liability and accountability,” the MPS submission states.
Mandatory training
While MPS described the draft guidance as “an important milestone”, it called on HIQA to ensure that it was workable for a variety of healthcare settings – including smaller community‑based services and dental practices – and not just hospitals.
MPS added that the guidance must be clear, concise, and easy to follow, to help busy healthcare professionals to easily identify the most important points.
The organisation also called for the guidance to set out the need for mandatory training on AI tools, saying that the current wording – that AI training “should be made available” – does not go far enough.
Diversity of settings
Dr Sarah Townley (deputy director at MPS) said that having national guidance would set a clear standard for safe, ethical use of AI in healthcare.
She added, however, that the document must reflect the diversity of settings covered by the scope of the guidance, and varying levels of knowledge on AI.
“Clinicians are already encountering new medico-legal challenges as AI becomes more deeply embedded in clinical practice,” Dr Townley stated, citing issues such as consent, accountability, and the exercise of clinical judgement.
MPS has created its own AI Safer Practice Framework, which is aimed at helping medical professionals to navigate these issues.