smartphone AI that spots eye disease: a threat for optical careers?

A smartphone that can screen for three serious eye conditions, with no internet required?

It sounds like a stretch, but a recent study found an offline AI doing exactly that, picking up diabetic retinopathy, glaucoma and AMD with real accuracy.

The important question, though, is where this kind of screening fits into eye care, and what it means for the people working in the optical industry.

the study, summarised in 20 seconds.

193 people. One smartphone-clip camera photographing the back of the eye. Two experienced eye specialists working blind, with no idea what the AI had already spotted.

The result? The tool matched the specialists strikingly well at flagging disease, and properly nailed glaucoma detection.

It’s offline. No cloud, no signal, no bulky tabletop kit. The AI does its thing on the device itself, in seconds, working in places a traditional fundus camera would never even reach.

some immediate limitations that come to mind.

While it’s true that the AI was very good at spotting disease overall, it was less reliable at catching early or mild cases of AMD and diabetic retinopathy – which means some problems could still be missed.

Also, this study was fairly small, run at a single site, and skipped any images that weren't clear enough to read – something a real screening programme out in the community can't do.

threat, tool, or opportunity?

"It's time to panic, there's a phone that now does my job" misreads what screening actually does.

This tool finds problems in people who weren't getting checked at all. It can't treat anyone. Every flag still needs a qualified professional to decide what happens next and explain it to the patient. More detection means more downstream work, not less.

But (yes there’s always a but) we'll be straight with you: nobody knows exactly what this looks like in five to ten years. Tools like this will change the job of an optometrist. Screening is more automated, and support staff can handle more of the imaging work, freeing optometrists up for the clinical calls.

Whether that's a net gain for the profession depends on how optical practices adapt, and that's a genuine unknown, not a reassurance we can just simply hand out.

the takeaway for people in the optical industry.

Nothing from this study fundamentally changes anything, for now at least. Yes, AI and technology keep getting better at spotting eye problems, and tools like this will likely find their way into more clinics and community screening programmes over the next few years. But spotting a problem and looking after the person who has it are two very different things. It still takes actual people, trained clinicians and support staff, to explain a result, decide what happens next, and care for the patient behind it.

If you're building a career in optometry, skills like these are worth picking up. If you're building an optical team, they're worth hiring for.

how we support you.

Whichever way the technology turns, one thing stays constant: opticians need skilled, adaptable people.

Talentshed's optical recruitment specialists help employers build teams ready for how the profession is actually changing, and help candidates find roles where their imaging and clinical skills are genuinely valued.

However the tools evolve, we're here to help you get the right people in place.

Talk to Talentshed today and see how we can support you.

Sources

• Ophthalmology Times – "Offline smartphone AI accurately detects DR, glaucoma, and AMD in real-world study" (2026).

• Primary Study: "Smartphone-based offline AI for multi-disease retinal screening: Real-world accuracy", European Journal of Ophthalmology (2026). Read the full study.