Search

FAQs

Commonly asked questions about the use of AI in healthcare and it’s regulation

Optometrist examining patients eyes at practice

1. What is the general status of the clinical utility of AI in healthcare?

There is frequent press coverage around the transformative prowess of AI in healthcare settings. Often hailed as the saviour of the NHS, sensationalised into headlines such as “Robots will replace doctors” or promised to release time for clinicians to care for patients through automation. The reality is not as clear cut, however. 

Regulation, NHS funding, patient trust and basic IT systems used across healthcare are likely to dampen the pace of change. While there are many potentially transformative innovations in development, at scale deployment will not happen overnight. Thankfully this offers opportunity for optometrists to better understand the most efficient AI products and how to best use them to improve patient care and workflow. The key is to educate yourself.

More information

Frontier AI: double-edged sword for public sector, Joseph Rowntree Foundation

More than just hype: how emerging AI use is assisting health and social care, The King's Fund

Machine Learning, Royal Society

NIHR A Horizon Scan of AI technologies in Healthcare

UCL Institute of Ophthalmology Inaugural Lecture with Professor Pearse Keane

Understanding AI in healthcare

2. What AI products are currently available to optometrists (from administrative through to clinical)? (section to be updated as regularly as possible)

The widespread use of optometry-specific AI is still in its infancy but there are innovations that could reshape optometry for the better in the near(ish) future. For example, the recent formation of Cascader, a partnership in UCL’s ophthalmology department and Moorfields Eye Hospital Trust has a mission is to build a suite of AI tools that enable earlier and more accurate detection of disease across all levels of eye care, including optometry. This initiative will be watched with interest in the coming months and years.

Worth noting, is that the UK's diabetic eye screening program is currently evaluating and beginning to implement Automated Retinal Image Analysis Systems (ARIAS) (these tools are registered AIaMDs), across some areas of the UK. With several systems showing efficacy at least equal to human gradersxxxiv, the widespread adoption of these tools in diabetic retinopathy screening could be soon.

More information

Application of Artificial Intelligence in Medical Imaging: Current Status and Future Directions, Yang, 2025, iRADIOLOGY, Wiley Online Library

Artificial intelligence in ophthalmology: opportunities, challenges, and ethical considerations, PubMed

Full article: Smart Devices in Optometry: Current and Future Perspectives to Clinical Optometry

Generative artificial intelligence in medical imaging: Current landscape, challenges, and future directions, He, 2025, Interdisciplinary Medicine, Wiley Online Library

Optician Online - Artificial intelligence 7: Language models for optometry

Wet AMD system at Moorfields

3. Who will be responsible if the AI goes wrong?

A frequent concern raised by healthcare professionals when using anything AI/automated is that patient harm will occur as a result of being given inaccurate advice or that some vital medical information is missed by the system. Thankfully, trust and accountability are key themes of emerging legislation.

For the moment, it will always be the practitioner who is responsible for decisions that affect the patient, AI should always be considered an assistive tool. For further guidance on the need to maintain your own responsibility for all clinical decisions, please look to the College of Optometrists’ Position Statement on AI.

More information

Artificial Intelligence and ethics, Royal Society

Ethics, Transparency and Accountability Framework for Automated decision-making

Inevitable challenges of autonomy: ethical concerns in personalized algorithmic decision-making, Humanities and Social Sciences Communications

10 things governments should know about responsible AI

4. What are governments doing to keep patients and public safe?

Looking at the global picture of how governments are responding to the advent of these technologies, many nations are various stages of legislation. Some offer comprehensive governance, many are developing focused legislation for specific use, with most stopping short at national AI strategies or policies alongside myriad voluntary guidelines and standards.xxxv The flagship EU AI Actxxxvi is a notable example of how legislation could work for the benefit of society, with risk and trust at the core, regulatory requirements being strict and specific for certain scenarios.

In the current absence of specific overarching legislation on AI, the UK has been trying to position itself as a world leader in the regulatory environment. This is demonstrated by the Medicines and Healthcare products Regulatory Agency (MHRA) working with the Artificial Intelligence/Machine Learning-enabled, International Medical Device Regulators Forum.xxxvii Trying to remain as agile as possible, the MHRA have produced policy principles on treating AI-powered products as Medical Devices under the existing framework.xxxviii Further, they have launched the AI Airlock working with industry and healthcare providers on pilots that sense test how regulation could and should work for various products that are likely to come to widespread use.

More information

Artificial Intelligence Act, Artificial Intelligence Act

Artificial Intelligence/Machine Learning-enabled, International Medical Device Regulators Forum

AI Airlock: the regulatory sandbox for AIaMD

Developing regulatory science to advance healthcare, UKRI

Global AI Law and Policy Tracker

Global AI law and policy trends update, IAPP

Implementation of the future regulations

MHRA emerging regulations for Medical Devices

6. I would like to participate in AI research, where would I find information?

References

xxxiv Automated retinal image analysis systems to triage for grading of diabetic retinopathy: a large-scale, open-label, national screening programme in England

xxxv Global AI Law and Policy Tracker

xxxvi EU AI Act: first regulation on artificial intelligence

xxxvii Artificial Intelligence/Machine Learning-enabled

xxxviii Impact of AI on the regulation of medical products