Government Accepts 44 Recommendations on Rules for AI in Healthcare

The UK government has accepted all 44 recommendations made by the National Commission into the Regulation of Artificial Intelligence in Healthcare, setting out a direction for how AI-enabled tools and medical devices should be assessed and monitored.

The response, published on 6 October 2026, follows the independent commission’s report issued in September. It supports a regulatory approach that accounts for how systems may change after deployment, rather than treating safety assessment as a single event at the start of a product’s life.

What could change

The commission’s recommendations cover areas such as medical-device classification, evidence of clinical performance, monitoring once products are used in practice and clearer responsibilities for developers, health providers and regulators. These proposals aim to allow useful technology to reach patients while maintaining safeguards.

The patient-safety questions

AI systems can produce errors, behave differently between populations or change after updates. The government’s response highlights the need for proportionate testing, transparency and routes to identify problems. Those questions are especially important when tools influence a clinical judgement or the information available to staff.

The next phase

Alongside the announcement, the Medicines and Healthcare products Regulatory Agency opened applications for the third phase of its AI Airlock regulatory sandbox, designed to test approaches to oversight of AI-enabled medical devices.

Accepting recommendations is the beginning of a policy and implementation process, not proof that new safeguards are already in force. Keep Updated UK will report subsequent guidance, rule changes and evidence of how the proposals work in practice.