Aipedia: Optometrists and opticians

Telling patients you use AI in eye examinations

Aipedia, by the Apex broking team · Last reviewed 2026-10-30

Patients should be told when AI helps assess their scans, images or symptoms, and told that the optometrist makes the decision. Transparency supports valid consent under GOC Standard 3 and is set to become an explicit College expectation.

Part of: AI in Optometrists and opticians

In short

GOC Standards 2 and 3 require effective communication and valid consent, and Standard 14 protects the confidentiality of images and data, including images shared online. The College of Optometrists' draft 2026 Guidance for Professional Practice adds transparency with patients as an expectation in its new AI section, and proposes a professional duty of candour for all registrants. GOC Standard 19.1 already requires openness when things go wrong, including AI-related errors. Where significant decisions on health data are made solely by automated means, section 80 of the Data (Use and Access) Act 2025 adds restrictions. Clear, short explanations and good records are the practical answer.

Why transparency matters

Patients may expect to know how decisions about them are made. In optometry, AI may analyse OCT scans, grade fundus photographs, suggest referral urgency or draft the record. A patient who later learns that software played a part, and was not told, may feel misled even if care was good. If something went wrong, lack of transparency can turn a clinical question into a trust problem and a complaint.

Transparency also protects the optometrist. Explaining that software assists but does not decide sets the right expectation: the patient knows a professional looked at their eyes. It also prompts the optometrist to be able to say, in plain words, what the tool does and what it does not do, which Standard 5.3 on keeping aware of digital technologies already expects.

The duties and expectations

The standards and guidance that shape what you tell patients are:

Standards 2 and 3: the GOC Standards of Practice require effective communication and valid consent. Telling patients about AI-assisted assessment supports both.
Standard 14: retinal images, OCT scans and records entered into AI or cloud tools must be kept confidential and secure. The GOC launch note for the 2025 standards flagged confidentiality of patient images shared online.
Standard 19.1: be open and honest with patients when things go wrong.
College guidance: the draft fifth edition of the Guidance for Professional Practice sets expectations on transparency with patients in a new AI section and proposes a professional duty of candour for all registrants. Final text is due in winter 2026.
Automated decisions: section 80 of the Data (Use and Access) Act 2025 restricts solely automated significant decisions on health data.

What to tell patients

Explanations do not need to be long. A practical approach:

Say that software helps analyse the scan or photo, and that the optometrist reviews the images and makes the decision.
Explain whether images are sent to an external or cloud service.
For recorded consultations, say so before recording starts and offer an alternative.
For chatbots and apps, make clear the patient is using an automated system and what it cannot do.
Answer questions honestly, including about the tool's limits.

Practice leaflets, website notices and privacy information help, but a short spoken explanation at the point of care is what most patients remember. Record that it was given and any objection.

Candour if an AI-related error occurs

Illustrative scenario: a practice discovers that its fundus grading tool was set to analyse only the central retina for several months, and that some patients' peripheral findings were not reviewed. The optometrist must decide what to tell affected patients. Standard 19.1 requires openness: patients should be told what happened, what it means for them and what is being done, including any recall for review. Saying "the software missed it" without accepting the practice's role is unlikely to satisfy a regulator or a patient.

An apology or explanation is not usually an admission of legal liability, but it is sensible to notify your indemnity provider promptly when an incident may lead to a claim, as policies commonly require early notification. GOC Standard 12.2 requires adequate indemnity, and cover usually responds subject to its terms.

Practical controls

You can reduce the risk with these steps:

Add AI use to patient information and privacy notices.
Use a short, consistent spoken explanation in the test room.
Record that patients were told and any objection.
Have a process for reviewing affected patients after a tool error.
Notify your insurer early when an incident may lead to a claim.

How Apex helps

Apex arranges PI for optometrists and optical practices. A broker can explain how your policy handles notifications and candour after an AI-related incident; see our optometrist PI guide.

Sources

Frequently asked

Do I have to tell patients that AI analysed their scan?

The GOC has no AI-specific rule, but Standards 2 and 3 on communication and consent support telling them, and the College's draft guidance makes transparency with patients an expectation. It is good practice to explain briefly and record it.

What if a patient objects to AI analysis?

Respect the objection where you can, record it and explain any effect on their care. You may be able to review the images without the AI tool. Where data would go to an external service, an objection is especially important to honour.

Does an apology after an AI error count as admitting liability?

An apology or explanation is not usually an admission of legal liability. Standard 19.1 requires openness when things go wrong. Notify your indemnity provider promptly, as policies commonly require early notification of incidents.

Is a privacy notice enough to tell patients about AI?

A privacy notice helps meet data protection duties, but few patients read it. A short spoken explanation at the point of care, recorded in the notes, does more to support valid consent under Standard 3 and the transparency the College's draft guidance expects.

Check your PI covers how you use AI

Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.

Start a PI quote Request a callback

Apex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.