Apple · Filed Jun 18, 2026 · Published Oct 29, 2026

Apple Files Patent to Nudge You Into Position for Eye-Scan Logins

Apple's patent gives an eye-scanning device a way to tell you what's wrong when it can't get a good look at you, so you can fix it instead of just failing.

An eye scanner tracks a user's eye and scans the iris pattern to confirm their identity for secure login. Drawing from patent filing US 2026/0340098 A1.
An eye scanner tracks a user's eye and scans the iris pattern to confirm their identity for secure login.
See all 22 drawings from this filing ↓
Publication number US 2026/0340098 A1 US20260340098A1
Applicant Apple Inc.
Filing date Jun 18, 2026
Publication date Oct 29, 2026
Inventors Martin Haller, Tom Sengelaub
US classification 713/186
Status when we published Waiting for an examiner (Jul 18, 2026)
Parent application is a Continuation of 18027919 (filed 2023-03-22)
Document 20 claims

What Apple's eye-scan alignment prompt does

Eye-based logins only work when the camera gets a clear look at your eye, and that is harder than it sounds. A headset can slip on your head, a lens can get smudged, and a clip-on lens can change the view, so image quality shifts from one attempt to the next. Apple's filing points out that image quality affects how well the scan works.

Apple's fix is for the device to check its own view first. If your eye is sitting at a bad angle, it tells you, with a buzz, a sound or a cue on screen, to look a different way or nudge the device. Then it tries the scan again.

The filing also describes a version where the camera itself tilts, moves or zooms, so you have less to do.

From the filing · CLAIM 21
… determine based on the evaluation that the current eye pose is not satisfactory for capturing images to be used in a biometric authentication process; signal the user to adjust their eye pose or to adjust a position of a device including the camera …

Translation: The system checks if your eyes are lined up correctly and tells you to move if they are not.

How the camera judges whether your eye is lined up

The claims cover a camera pointed at your eye region plus a controller (the chip and software that run the logic). The controller first works out your eye pose, meaning where your eye sits and which way it points relative to the camera. It then decides whether that pose is good enough to capture images for a biometric scan (identifying you from your body).

If the pose falls short, the controller signals you to change your gaze or move the device, then runs the scan on images taken after you adjust. The claims also cover repeating the check and the prompt until the pose is satisfactory. To judge a pose, the controller can test the images against objective criteria, such as:

  • Gaze direction, exposure and contrast
  • Shadows, streaks and unwanted reflections
  • Sharpness and evenness of lighting
  • Anything blocking the eye

Once it has good images, it can score different features (iris patterns, eye veins, eyelids, eyebrows, skin or nose features) and use the best ones for the match. The hardware can be infrared LEDs with an infrared camera, in a headset, a handheld device or a wall-mounted unit.

The description adds a variation: the controller can direct the camera to tilt, move or zoom, and the feedback to you can be a vibration, a sound or an on-screen indication.

From the filing · THE ABSTRACT
… the eye pose may be improved by the user manually adjusting the device or their pose/gaze direction in response to a signal from the controller, and/or in which the imaging system is mechanically adjusted …

Translation: The device tells you to reposition yourself or automatically moves its camera to get a better scan.

Why failed eye scans on headsets could get rarer

If you have ever had a face or fingerprint scan fail for no clear reason, you know the frustration of not knowing what to fix. This filing adds a hint: the device tells you what to change. The most obvious fit is a headset, where the description says cameras often sit at awkward angles and the device can shift on your head, but the claims also name handheld and wall-mounted devices.

For Apple the appeal is reliability. If scans work on the first or second try, people have more reason to use eye-based sign-in. The filing's own tradeoff diagram notes that automating the fix, such as shifting the camera, can make setup easier on you. A filing is not a product, and nothing here says this ships.

That makes this Apple's 47th filing we've tracked since May on our eye and hand controls watchlist, adding to earlier work on flat apps reading gaze and touch and cameras in the nose bridge.

Editorial take

The cost here is your patience. A device that stops to say your eye is out of position adds a step to something that should feel instant, and the claims let the check and the prompt repeat until the view is good enough. If your headset sits loose or a lens is smudged, you could get nagged more than once.

The filing admits a second trade. A more complex machine can ask less of you, for example by tilting the camera for you. That means extra moving parts in a small headset, which adds cost and one more piece that can wear out.

Even so, the trade reads as worth it. A clear instruction like "look a bit to the left" beats a failed scan with no explanation, and the prompt costs only a moment when it works. The open question is how often the device would speak up, and the filing leaves that unanswered.

Get our take in your Top Stories

Liked this breakdown? Add Patentlyze as a preferred source on Google, and our plain-English take shows up more often in your Top Stories the next time Apple patent news breaks.

There are more where this came from

We read every patent application Big Tech publishes and send you the ones worth knowing. Plain English, free, every week.

The drawings

22 drawing sheets from US 2026/0340098 A1 · click any drawing to enlarge

Two ways cameras capture views of your eye, drawing sheet 1 from US 2026/0340098 A1
Sheet 1 Two ways cameras capture views of your eye
Scanning through a lens using a ring of lights, drawing sheet 2 from US 2026/0340098 A1
Sheet 2 Scanning through a lens using a ring of lights
Balancing hardware complexity against user effort during setup, drawing sheet 3 from US 2026/0340098 A1
Sheet 3 Balancing hardware complexity against user effort during setup
Stored lighting presets selected to illuminate your eyes properly, drawing sheet 4 from US 2026/0340098 A1
Sheet 4 Stored lighting presets selected to illuminate your eyes properly
How lighting adjusts until scans become clear, drawing sheet 5 from US 2026/0340098 A1
Sheet 5 How lighting adjusts until scans become clear
Reading facial features around the eye alongside iris details, drawing sheet 6 from US 2026/0340098 A1
Sheet 6 Reading facial features around the eye alongside iris details
Choosing the best biometric details to verify your identity, drawing sheet 7 from US 2026/0340098 A1
Sheet 7 Choosing the best biometric details to verify your identity
Multiple cameras viewing your face from different angles simultaneously, drawing sheet 8 from US 2026/0340098 A1
Sheet 8 Multiple cameras viewing your face from different angles simultaneously
Show the other 14 sheets
Choosing or combining multi-angle scans for eye recognition, drawing sheet 9 from US 2026/0340098 A1
Sheet 9 Choosing or combining multi-angle scans for eye recognition
Adding extra lenses to improve the scanner's view, drawing sheet 10 from US 2026/0340098 A1
Sheet 10 Adding extra lenses to improve the scanner's view
Correcting distorted eye images caused by extra lens layers, drawing sheet 11 from US 2026/0340098 A1
Sheet 11 Correcting distorted eye images caused by extra lens layers
Bending reflected light to give the camera a clearer angle, drawing sheet 12 from US 2026/0340098 A1
Sheet 12 Bending reflected light to give the camera a clearer angle
Flashing different light colors one by one to reveal details, drawing sheet 13 from US 2026/0340098 A1
Sheet 13 Flashing different light colors one by one to reveal details
Splitting sensor pixels to capture two light wavelengths at once, drawing sheet 14 from US 2026/0340098 A1
Sheet 14 Splitting sensor pixels to capture two light wavelengths at once
Taking turns scanning the eye with different colors of light, drawing sheet 15 from US 2026/0340098 A1
Sheet 15 Taking turns scanning the eye with different colors of light
Recording multiple light wavelengths in a single shot, drawing sheet 16 from US 2026/0340098 A1
Sheet 16 Recording multiple light wavelengths in a single shot
Sending feedback to line up the eye sensor correctly, drawing sheet 17 from US 2026/0340098 A1
Sheet 17 Sending feedback to line up the eye sensor correctly
Prompts telling you to adjust your gaze for a scan, drawing sheet 18 from US 2026/0340098 A1
Sheet 18 Prompts telling you to adjust your gaze for a scan
Moving the internal camera automatically to get a clear scan, drawing sheet 19 from US 2026/0340098 A1
Sheet 19 Moving the internal camera automatically to get a clear scan
Phone layout with an eye camera and ring of lights, drawing sheet 20 from US 2026/0340098 A1
Sheet 20 Phone layout with an eye camera and ring of lights
Headset layout tracking your gaze to verify your identity, drawing sheet 21 from US 2026/0340098 A1
Sheet 21 Headset layout tracking your gaze to verify your identity
Dual-camera setup scanning both eyes at the same time, drawing sheet 22 from US 2026/0340098 A1
Sheet 22 Dual-camera setup scanning both eyes at the same time
Browse all the drawings 22 drawing sheets
Patent filing page

Source. Full patent text and figures from the official USPTO publication PDF.
Reader comments

Be the first to weigh in

Start the discussion

Real name or a handle, either is fine. Comments are read by a person before they appear, so allow a little time. Keep it about the filing.