On-device camera versus Bluetooth proximity: two ways to auto-lock a Mac.

The two main approaches

When you leave your desk, you want your Mac to lock automatically before someone else can use it. But how should it detect that you have gone? The answer depends on which approach you use: either reading the built-in camera to check whether your face is present, or checking Bluetooth proximity to a phone or watch you are carrying.

These are not the only possible signals (keyboard and mouse idleness, location-based geofencing, and others exist), but they are the most common in real Mac apps available today. Each has a distinct set of strengths and weaknesses.

The key is understanding what you are trading off. Neither approach is universally superior; the better choice depends on your environment, habits, and what you value most.

Bluetooth proximity approach

The Bluetooth proximity method pairs your Mac with a device you carry, typically an iPhone or Apple Watch. The Mac periodically checks the signal strength (RSSI, or received signal strength indicator) of your device over Bluetooth. When the signal drops below a threshold (indicating you have walked out of range), the Mac locks.

Real apps using this method include Near Lock and Tether Lock, which lock and unlock based on your iPhone's proximity. The logic is straightforward: your phone is with you, so if the Bluetooth signal fades, you must have left. It is the same mechanism behind Windows Dynamic Lock, which the Mac has no built-in equivalent of.

Strengths of Bluetooth proximity

• No camera permission required. Some users are uncomfortable granting camera access, even for local processing. Bluetooth proximity avoids this entirely.
• Works in any lighting. Bluetooth operates independent of lighting conditions. Darkness, backlighting, or a covered camera do not affect detection.
• Works with any Mac. Every Mac with Bluetooth can support this approach. You do not need a camera, external or built-in.
• Simple logic. The mechanism is easy to understand: if your phone's signal goes away, you are away.

Limitations of Bluetooth proximity

• Requires a paired device in your pocket. The Bluetooth device must travel with you. If you leave your iPhone on your desk while you step away, the lock will not trigger. If your watch battery runs out, the system fails.
• Bluetooth range limitations. Typical Bluetooth range is 20 to 100 metres depending on conditions, walls, and interference. If you are in the same building but outside Bluetooth range, the lock works as intended. But if you are briefly in another room, or if you work in a dense office with many overlapping signals, false locks can occur.
• Battery dependency. Your iPhone or Apple Watch must be charged. A dead watch battery, or a phone that runs out of power, silently breaks the lock system. You may not realise until you return to your desk.
• Latency and glitches. Bluetooth connections can be unstable. A temporary signal dropout (a glitch that lasts a few seconds) might trigger a false lock. This is why real proximity apps implement timeout logic to avoid locking on every brief signal blip, but tuning this delay is a balancing act.
• False positives with shared spaces. If someone else is nearby with a similar Bluetooth device, or if your device is near the edge of range, the system may make the wrong decision.

Camera-based face presence approach

The camera approach uses your Mac's built-in (or external USB/Thunderbolt) camera to detect whether a face is present in the frame. Using Apple's Vision framework, the system performs on-device face presence detection: it checks whether pixels in the frame match the pattern of a human face, without uploading the image or storing photographs. When the camera stops seeing a face for a configured duration, the Mac locks.

Peeku is the primary example of this approach on macOS. It runs face detection continuously in the background and locks when the camera detects your absence.

Strengths of camera presence detection

• Direct physical detection. The system sees whether you are actually sitting in front of your Mac. If you step away, even with your phone still in Bluetooth range, the lock will trigger.
• No battery dependency on external devices. The detection runs on your Mac itself. There is no reliance on your phone's battery or watch charge. The only battery involved is your Mac's, which you are already monitoring anyway.
• No external device required. If you do not want to carry a phone or wear a watch, face detection still works. It can optionally use an Apple Watch as a reinforcement signal, but the Watch is not required for locking to work.
• Fewer false locks from range glitches. Because the detection is based on your physical presence rather than a Bluetooth signal, there is no ambiguity from connectivity dropouts or devices drifting in and out of range.

Limitations of camera presence

• Requires camera permission. macOS will prompt you to allow camera access. Some users decline on principle. Additionally, the camera cannot be covered or disabled without affecting the lock system.
• Lighting matters. Face detection works best in reasonable lighting. If your desk is backlit, dimly lit, or in shadow, the system may struggle. Sunglare directly on the camera lens can also reduce confidence.
• Requires line of sight. The camera must have a clear view of your face. If you work with a monitor that completely obscures your face from the built-in camera, or if you sit to the side rather than facing the screen, detection may be unreliable.
• Not all Macs have cameras. Older Mac desktops, or external monitors, may not have a camera. You can attach a USB or Thunderbolt camera, but that is an extra step.
• Brief look-aways trigger caution. If you glance down at a notebook, lean back briefly, or turn your head, the system may momentarily lose face detection. Real implementations add a grace period (a few seconds of face absence before locking) to avoid locking mid-task. This grace period needs tuning: too short and you lock when reading; too long and a visitor can unlock your screen if they know how long to wait.

Comparison table

Hybrid approaches

Some systems use both signals at once. Peeku, for instance, uses face detection as the primary signal and can optionally pair with an Apple Watch via Bluetooth. The Watch acts as a gating signal: both the camera and the Watch must agree you are away before the lock engages. This reduces false locks from either mechanism alone.

If the camera falsely detects absence (due to poor lighting or you stepping out of frame briefly), the Watch signal might prevent the lock. Conversely, if the Watch drifts out of range momentarily, the camera can confirm you are still present. The combination is more robust than either alone, at the cost of added complexity.

Special case: touch-based unlock (MacID)

One other approach exists on the Mac: MacID uses your built-in Touch ID to unlock your Mac, triggered by user action. This is different from walk-away detection; it is not "automatic" absence detection but rather "user-initiated unlock with biometric confirmation." It does not lock automatically when you leave, but it does avoid the need for a password when you return. This is a useful complement to any auto-lock system but not a replacement for walk-away detection.

How to choose

Your choice depends on your specific situation. Ask yourself these questions:

• Do you always carry your phone or wear a watch? If not, Bluetooth proximity will miss the moment you leave. If yes, it is reliable for your pattern.
• Is your desk well-lit and does your camera have a clear line of sight to your face? If yes, camera detection is a strong choice. If your desk is dark or you often sit away from the camera, it may struggle.
• Are you comfortable granting camera access? If not, Bluetooth proximity is the better fit. If you do not mind, camera detection offers more independence from external devices.
• Do you frequently leave your phone on your desk by mistake? If so, camera detection catches this case while Bluetooth does not.
• Do you want the simplest possible setup? Bluetooth proximity is conceptually simpler. Camera detection requires more tuning of detection sensitivity and grace periods.
• Is battery redundancy important? If you want a system that works even when your phone dies, camera detection is more reliable.

Beyond auto-lock: wellness coaching

An interesting side effect of running face detection is that you can also offer wellness features. Apps like Peeku use the camera to detect posture, eye openness, and gaze direction, enabling coaching for breaks, walk reminders, and posture hints. Bluetooth-based systems cannot easily offer these features because they have no visibility into what you are actually doing; they only know whether a device is in range.

If wellness coaching is important to you, a camera-based system offers more potential. If you only care about locking, Bluetooth proximity is sufficient.

Privacy considerations

Both approaches process data locally on your Mac. Neither requires creating an account or uploading raw camera frames or Bluetooth signals to a company server. However, they differ subtly:

• Bluetooth proximity: The app knows when your device is in range and out of range. This is a temporal record of your presence. If an app logged this and sent it to a server, it could build a picture of your schedule. Real proximity apps (Near Lock, Tether Lock) do not do this by default, but the signal itself is sensitive.
• Camera face presence: The app processes your camera frames locally, never stores or uploads photos, and only records the binary decision "face present / absent". If it uses feature prints to distinguish your face, those are mathematical vectors (not photos) stored on your Mac, not cloud-synced in the baseline version. Again, if an app sent this history to a server, it could track your schedule. Peeku does not do this by default, but the potential exists.

In both cases, the privacy claim rests on the app itself being trustworthy. Choose an app from a developer with a clear privacy policy and no incentive to monetise your data.

Conclusion

Camera-based face presence and Bluetooth proximity are complementary solutions to the same problem. Camera detection is better if you want independence from external devices and can offer richer coaching features. Bluetooth proximity is better if you are uncomfortable with camera access, have inconsistent lighting, or strongly prefer a simple system. Many users find a hybrid approach (camera with optional Watch reinforcement) strikes the right balance.

The best choice is the one that matches your workflow, your desk setup, and what you value most in a security and wellness system.