Headlines about AI taking over clinical professions are not new, and eyecare has had its own version of this conversation for a while now: will AI replace opticians and optometrists? Retinal screening tools can flag disease markers in seconds; virtual try-on lets people choose frames without setting foot in a practice; scribes draft exam notes while the optometrist is still talking to the patient.
Taken together, it is easy to see why the question keeps surfacing: is this the beginning of the end for opticians and optometrists?
I have heard this worry before. In 1992 I started on a front desk with a paper appointment book, typewritten schedules, handwritten recall postcards, and a filing system that filled whole rooms, and that practice ran on paper until 2010. When electronic medical records became commonly used, the same fears came with them: that screens would come between us and our patients, and that the people on the desk and in the testing room would be designed out.
That is not what happened. Those technologies opened access to care in rural communities and gave us the population health improvements we can see today, in our own lifetime. I believe AI can make even bigger leaps for patient care, and that eyecare can be at the leading edge of it. So, the short answer is no, and the profession’s own bodies agree.
AI is changing how eyecare practices operate, but the roles built on clinical judgement, hands-on skill, and patient trust are not the parts of the job that AI is positioned to take over. This article looks at why the fear persists, what AI is genuinely capable of right now, and where the optometrist and the dispensing optician remain the people patients actually need in the room.
The Question the Trade Press Keeps Asking
Part of the reason this topic keeps coming up is that AI adoption in eyecare has moved quickly from pilot projects to everyday tools. Practices are using AI for OCT and retinal image analysis, for automated patient recalls, for chat-based booking, and for administrative work that used to take up a fair portion of a clinician’s and practice’s day.
Optometry Today, the journal of the Association of Optometrists, has been running this conversation for months, most recently reporting on a University of Leicester survey of UK optometrists’ trust in AI tools.
When trade publications ask hard questions about AI, and professional bodies publish formal positions on it, that is a good thing. It means new technology gets tested before it becomes standard practice.
So, when we ask whether AI will replace opticians and optometrists, the answer isn’t coming from marketing copy or guesswork. It comes from the same bodies that set clinical standards for the profession.
What AI is Actually Good at Today
The AI tools already in use across UK and international practices fall into a few clear groups: documentation, screening support, image analysis and day-to-day operations.
AI scribes can draft exam notes, referral letters, and claim forms while the clinician is still talking to the patient. The optometrist still checks everything and signs off before it becomes part of the record.
Image analysis tools help spot patterns linked to conditions like diabetic retinopathy, glaucoma, and age-related macular degeneration in OCT scans and retinal photos, flagging them for a clinician to look at.
On the operations side, AI-assisted recall and reminder systems help bring patients back for their regular eye examinations on time. This matters because a missed recall means a condition goes unwatched for another year.
That last group is where the admin gains show up most clearly, and it is the part of the job I know best. On a paper front desk, recall meant writing postcards by hand and working through phone lists.
Automated, multi-channel reminders can win back much of the time and revenue that no-shows and late cancellations cost every month, and free the team to focus on the patients actually in front of them instead of chasing confirmations by phone.
None of this replaces the person making the clinical call. It just clears away the administrative work around the patient visit, so the clinician spends less time typing, chasing, and re-entering data, and more time on the exam itself.
Bigger Than One Practice: AI and Population Health
The move from paper to electronic records did more than tidy up the front desk. It made it possible to see patterns across thousands of patients, to reach people in rural communities who were a long way from the nearest practice, and to follow up the ones who would otherwise have fallen through the cracks between visits.
AI could help take that further. An eye examination is one of the few routine appointments where signs of diabetes, high blood pressure and other systemic conditions can be seen directly. When AI helps practices recall the right patients at the right time, flag the scans that need a closer look, and connect findings to the wider care team, eyecare stops being a single appointment and becomes part of how a population stays healthy. That is why I think eyecare is a very important part of overall healthcare.
Where Clinical and Dispensing Skill Still Lead
An AI model can flag a suspicious retinal image. It can’t sit with a patient and explain what that finding means for their driving licence, their diabetes management, or their fear of losing their sight. It can’t weigh a patient’s full history, their lifestyle, their other medications, and their own description of their symptoms, and turn all of that into a recommendation and plan of action that fits the patients’ lifestyle.
This is the point that comes up in almost every serious discussion of AI in eyecare: these tools are good at spotting patterns in a narrow, well-defined task. Clinical practice rarely stays narrow or well-defined.
What a patient presents with is often unclear, with mixed signals or context that only comes out through conversation and a hands-on exam. Refraction still needs judgement calls about a patient’s comfort and visual needs that an autorefractor reading alone can’t give you. Diagnosis often means ruling things out through direct examination, not just feeding an image into a model.
The same is true in the dispensary. Measuring accurately for progressives, fitting and adjusting a frame so it sits correctly on a particular face, helping someone through their first weeks in varifocals, and repairing a pair of glasses so a patient can see on the drive home are physical, skilled tasks. Virtual try-on can help someone narrow down a style, but it cannot adjust a bridge, set the pantoscopic tilt, modify temple lengths or adjustments or notice that a lifestyle question changes the right lens for that patient. A good fit still happens in person.
Then there’s accountability. When an AI tool makes a recommendation, someone still has to decide whether to act on it, and that someone is a registered professional whose name and judgement are on the record. AI can support that decision. It can’t make it, and regulation doesn’t allow it to either.
What UK Regulators and Professional Bodies Are Saying
In June 2025 the College of Optometrists published the sector’s interim position on AI in eyecare, put together with national optical bodies including the Association of Optometrists (AOP) and Association for Eye Care Providers (FODO). The position is clear: AI is there to support clinicians, not replace them. It also calls for robust testing, transparency, and clinician training before AI tools become a normal part of everyday practice.
The profession is also looking into how practitioners feel about this. A survey developed by the University of Leicester and OasisEye explored trust, perceived clinical usefulness, and how confident practitioners feel relying on tools that have often rolled out faster than the evidence behind them.
The Patient Relationship is Still Human
Ask most patients what they actually want from a visit to their optician and it goes well beyond a correct prescription. They want someone to explain what a finding means for them specifically, to notice when something feels off even before it shows up on a scan, and to be reassured by a person who has seen hundreds of similar cases and can speak to their situation with genuine confidence.
That relationship is part of why recall and follow-up matter so much clinically, not just commercially. A patient who trusts their optician is more likely to come back for the next exam, more likely to follow through on a referral, and more likely to mention a symptom they might otherwise dismiss as nothing. AI can help make sure that patient gets reminded to book. It cannot build the relationship that makes them want to show up.
What This Means for Your Practice
For practices weighing up how much to invest in AI, the useful framing is not replacement versus resistance. It is where AI genuinely reduces admin load and where the clinical, dispensing and relational work still needs a person.
It is also a people project as much as a technology one. I saw that with the move to electronic records: the practices that did best brought their teams along, explained what was changing and why, and gave the front-desk team room to grow into patient experience roles as the paperwork shrank. The same will be true with AI.
Administrative tasks such as recall, reminders, note drafting, and basic scheduling are strong candidates for AI-assisted automation, since the upside is largely about time saved rather than clinical decisions being made. Screening and triage support is worth adopting carefully, with proper validation and clear rules about clinician sign-off. Resposibility for diagnosis, treatment planning, dispensing, and anything involving a conversation with the patient about their care should stay firmly with the optometrist or dispensing optician.
Getting this balance right often comes down to the systems a practice is already running on. Choosing practice management software built specifically for eyecare, rather than a generic system with optical features bolted on, makes it far easier to bring in AI-assisted tools without fragmenting the patient record or creating extra manual work.
A practice running on an integrated platform such as Acuitas 3 already has recall, communication, and patient records in one place, which can provide a strong foundation for adding AI-assisted tools.
The practices that will benefit most from AI over the next few years are unlikely to be the ones chasing every new tool. They will be the ones who used AI to clear administrative bottlenecks so their optometrists and dispensing opticians could spend more time doing the part of the job that patients actually come in for.
I started with a paper appointment book and a room full of files. I have watched technology widen access to care and improve the health of whole populations within a single career. AI is the next, and bigger, step, and I want eyecare to be the profession that shows the rest of healthcare how to do it well.
