The recent decision by NHS England to pause its most ambitious AI initiative to date - an effort to apply predictive models across the health records of 57 million people - has raised serious questions. Yet its suspension, following concerns about patient consent and transparency, signals a broader issue we can no longer ignore. As a colorectal surgeon and someone closely involved in the AI space, I see this not as a failure, but as a signal - one that reminds us we can’t afford to treat governance as an afterthought.
The promise of AI in surgical care is already being realised in small but meaningful ways. I’ve witnessed the benefits in areas such as robotics, polyp recognition during endoscopy, enhanced operative planning, and even patient flow management. These aren’t abstract possibilities - they are real, current tools helping us work more effectively and safely. But even the most advanced system will fall short if we don’t first earn and keep the trust of the people we serve. What’s missing right now is a national, clinically informed approach to how we introduce and scale AI across the NHS.
We've seen huge enthusiasm, and significant financial backing, but the governance structures haven’t kept pace. That imbalance is precisely what this pause reflects. In my view, several steps are urgently needed:-
- A strong, transparent governance framework that ensures the use of AI respects patient autonomy and data privacy, and is built around clear, consent-based principles
- Active involvement of clinicians from the outset - AI must be designed for use in real clinical environments, with genuine input from those who will depend on it in daily practice
- Support and guidance for local systems, so that hospitals and Integrated Care Systems are not left to interpret vague national ambitions without practical tools or safeguards
This isn't a reason to pull back. It’s an opportunity to get it right. We can either build systems that align with clinical reality and public expectation - or we can continue to see projects stall due to avoidable oversights. AI in surgery isn’t the problem. Poor planning is. And it’s our responsibility - as clinicians, as leaders, and as stewards of patient trust - to make sure that planning improves.
Kiran Altaf
Director of Inclusivity and Diversity
