The integration of Artificial Intelligence (AI) into the National Health Service (NHS) is more a question of “when and how”, rather than “if”. From identifying tumours on chest X-rays to summarising complex patient consultations, AI tools are designed to streamline efficiency and support diagnostic accuracy.

However, a recent report from the Medical Protection Society (MPS) seeks to highlight that law reform is needed to accommodate these developments in technology because doctors could be held liable for clinical negligence if an AI tool makes a catastrophic error. The MPS describes doctors becoming a “liability sink” for technological mistakes. Are those fears well founded?

AI concerns raised by MPS

The central concern raised by the MPS is that the current legal framework is or will struggle to keep pace with the speed of technological change. If an AI tool suggests a diagnosis, or fails to detect a malignancy, who is ultimately responsible for that error?

Currently, the law in England and Wales generally places the burden of clinical responsibility squarely on the shoulders of the treating clinician. The MPS argues that unless the law is overhauled clincians may find themselves unjustly liable for decisions influenced, or even dictated, by algorithms they did not design.

I don’t accept those fears as likely to materialise, as I hope this blog will show. In addition, any reform to accommodate the MPS’s concerns, risk negligent care being absolved it if arose at the hands of AI. From a claimant’s perspective, we must be careful not to create a legal landscape where accountability is diluted.

AI and the role of medical professionals

Healthcare is not a tick-box operation. Assessing the standard of care requires that a clinician acts in accordance with a responsible body of medical opinion. If a doctor relies on an AI tool that misses a diagnosis, the legal question is not "was the AI wrong?" but "did the doctor act reasonably in relying on that tool?". Crucially that assessment will be based on what other doctors, (their clinical peers in essence) consider to be acceptable.

Clinical negligence is, fundamentally, about the failure to provide a reasonable standard of care. Doctors cannot absolve themselves of their professional duties by deferring to technology. An algorithm does not have a duty of care to a patient; a medical professional does.

If we allow a system where clinicians can deflect blame entirely onto software, we risk creating an "accountability vacuum." Patient safety relies on human judgement and oversight. If a tool suggests a treatment plan, the human clinician must apply their own professional judgment to verify that advice. Medical practice evolves, changes and varies as time progresses and the legal framework is set out to adequately deal with and accommodate that. The fears expressed by the MPS and the calls for reform I would respectfully suggest are misplaced.

AI and patient safety

The introduction of new tools into the NHS should never lower the bar for patient care. In fact, it should be the opposite.

We have seen this evolution before. When new surgical techniques or diagnostic methods are introduced, the law has adapted to define the reasonable standard of care in that new context. The same must apply to AI. If a Trust introduces an AI tool for screening, the staff using it must be properly trained, and there must be clear protocols for how that tool’s output should be utilised and checked.

If a doctor follows a flawed AI recommendation without critical assessment, that is not an issue of "technological liability"—that is an issue of clinical oversight. We must ensure that innovation does not lead to the erosion of patient safety or indeed good training for doctors. If doctors can rely on AI without clinical judgement, then surely they are superfluous. That of course is simply incorrect and a doctor’s knowledge and skill are paramount to a functioning healthcare system.

Future innovation

Innovation in the NHS is welcome, but it must be accompanied by robust governance and judgment. For patients, the fundamental protection remains the same: the expectation of high-quality, human-led clinical judgment.

Technology should be a tool to empower doctors to provide better care, not a shield to deflect responsibility. As these tools become more prevalent, the focus should still be on patient safety and delivering good care; innovation will not be utilised to deflect blame, even as the law stands now.