UK Doctors Could Face Claims Over NHS AI Errors

Why It Matters

Clarifying AI liability safeguards clinicians and ensures patients can seek redress, while addressing power constraints is essential for scaling AI infrastructure without overloading the grid.

Key Takeaways

• MPS warns doctors could become liability sink for AI errors
• Calls for AI tools to be treated as products under 1987 Act
• NHS reviewing liability guidance; NHS Resolution drafting AI rules
• US data centers may consume up to 12% electricity by 2028
• Grid upgrades and firm power needed for AI campus expansion

Pulse Analysis

The rise of artificial intelligence in clinical settings has outpaced the legal framework governing medical negligence in the United Kingdom. The Medical Protection Society’s recent alert highlights a gap: doctors may be held personally liable for errors generated by third‑party AI systems, even when they merely follow the tool’s recommendation. By advocating for AI to be classified as a product under the Consumer Protection Act 1987, the MPS seeks to shift responsibility toward developers and manufacturers, mirroring product liability regimes in other sectors. This approach could protect clinicians from becoming de‑facto liability sinks while preserving patients’ right to compensation when technology fails.

Policy makers are already responding. The Department of Health and Social Care has pledged a review of the MPS’s recommendations, and NHS Resolution is drafting new guidance on AI‑related negligence claims. The emerging consensus emphasizes rigorous documentation of AI involvement and clear oversight protocols, ensuring that any AI‑driven decision is traceable and that clinicians retain ultimate accountability. Such safeguards aim to balance innovation with patient safety, encouraging broader AI adoption without exposing healthcare providers to undue legal risk.

Across the Atlantic, the rapid expansion of AI workloads is creating a parallel infrastructure challenge. U.S. data centers, responsible for roughly 4.4% of national electricity in 2023, could account for up to 12% by 2028, according to a DOE‑backed study. This surge strains local transmission networks, where capacity upgrades often lag behind demand. Companies are exploring flexible workload placement and negotiating firm power contracts, but the long‑term solution will require coordinated investment in grid reinforcement, renewable generation, and even nuclear projects. Addressing power reliability now is critical to sustaining AI growth without compromising grid stability or escalating energy costs.