Apple Wants To Eliminate The Camera While Meta Fills The LED Gap: Privacy Is Reshaping AI Glasses Design

Online reports suggest that Apple’s "N50" smart glasses might drop the camera feature, opting instead to function solely as AI-assisted sensory devices.

This is not an isolated incident; in July 2026, Meta rolled out a software update for its smart glasses that immediately disables the camera if the system detects the white LED indicator light next to it has been tampered with—whether by drilling, covering it with nail polish or tape, or manipulating the software.

This patch is far from superfluous, as it addresses highly sensitive privacy concerns. The rumored decision by Apple to scrap the camera on its smart glasses is an extension of this same issue.

Privacy concerns are actively shaping the hardware design of AI glasses.

Google Glass establishes the rule: "An LED must be present."

When Google released Google Glass in 2012, it went down in history as one of the company's biggest failures; the device suffered from numerous flaws, with privacy concerns being the most significant.

The glasses allowed users to take photos via voice commands or by pressing a button on the top, yet they lacked an LED indicator light, leading many to view them as a tool for covert surveillance. Users wearing the device faced ejection by security guards in movie theaters and were denied entry to restaurants. By 2015, Google was forced to completely abandon the consumer version of Google Glass.

Despite its failure, Google Glass established a golden rule for the industry: AI glasses must include a visual indicator when recording. Since then, the LED light has ceased to be an optional accessory and has become an essential, standard feature. All AI glasses equipped with cameras—such as Meta Ray-Ban, Rokid, and RayNeo—now include an LED indicator light.

LEDs have been compromised, driving the need for "anti-tampering" and "physical isolation" measures.

The LED indicator became a standard feature, yet it was easily bypassed. An engineer named Bong Kim prompted Meta to release a software patch after posting a YouTube video demonstrating how to disable the LED circuit using a small drill bit; although the light was permanently extinguished, the system continued to register it as functioning normally. Vendors also introduced light-blocking stickers to easily obscure the LED's glow. These modifications led to serious consequences; for instance, modified smart glasses were used to secretly film flight attendants, sparking public outcry. Beyond user-driven issues, Meta’s own actions ignited even greater controversy. A 2024 investigation by Swedish media revealed that users had captured private footage—including images of bank cards and other sensitive data—using Ray-Ban Meta glasses, which was then sent to a data outsourcing firm in Kenya for review. The revelation triggered a joint protest from over 70 U.S. civil rights organizations, though Meta maintained that the practice complied with its privacy policy regarding human review. Meanwhile, Wired reporters discovered dormant facial recognition code within the Ray-Ban Meta companion app; this code allowed the glasses to capture and cross-reference facial images of people encountered while the wearer was walking. Although Meta’s Vice President of Communications, Andy Stone, stated that this was merely a pilot project rather than a fully implemented feature, consumers remained unconvinced and quickly filed a class-action lawsuit, alleging that Meta’s advertising claims of the product being "designed with privacy in mind" were deceptive.

The adverse effects resulting from these incidents have also compelled designers to make continuous improvements.

Standard LED lights are no longer sufficient to meet the need for protecting user privacy; consequently, Meta has implemented a hardware-level measure that escalates the response to unauthorized modifications from mere warnings to a complete hardware lockout. In response, many developers are revising their designs—such as the Samsung Galaxy Glasses, which feature a "dual privacy indicator" setup. China Industry News notes that mainstream smart glasses have widely adopted protective measures like indicator lights and physical switches, marking a shift in design philosophy from a matter of "trust" to one of "physical separation."

Apple's solution

Apple’s approach differs significantly from Meta’s. According to Bloomberg, there are three design options for Apple’s smart glasses—internally codenamed N50—regarding the camera. The first option involves completely omitting the camera and relying solely on voice interaction and AI assistant features; while this avoids the ethical risks associated with covert recording, it sacrifices the ability to visually perceive the physical world. The second option retains the camera but prohibits it from taking photos or videos; instead, the captured footage is used exclusively for AI analysis and is immediately discarded afterward. The third option is similar to the Meta Ray-Ban glasses, supporting both camera and AI capabilities.

If the hardware-level protections of a camera are bound to be compromised or spark controversy eventually, why not eliminate the risks at the source?

This approach aligns with Apple's privacy-by-design philosophy. Apple Intelligence employs an on-device-first architecture where the majority of AI processing occurs locally, keeping much of the relevant data on the phone. When cloud processing is required, the system prioritizes "Private Cloud" transmission with encryption; data is used solely to fulfill the immediate request and is destroyed afterward, ensuring that even Apple cannot access user data. Meta’s privacy policy differs, adopting a "collect first, remediate later" strategy—cameras remain active continuously, with safeguards and updates applied only after issues arise. In contrast, Apple chooses to eliminate the source of risk at the architectural level; for instance, if on-device AI can handle environmental awareness, there is no need to store raw footage, effectively transforming the camera from a traditional image-capturing device into a mere sensor.

Media reports indicate that Meta is making improvements, having compressed a multimodal AI model to under 2 GB; this eliminates the need to upload raw image data to the cloud while using the glasses. If this technical approach proves viable, the concept of "camera-less" design might give way to "storage-less" design. Moving away from cameras is unlikely to become mainstream anytime soon; Meta’s sales of over 7 million pairs of smart glasses demonstrate that photography remains a primary driver for consumer purchases of AI eyewear. Currently, there are two approaches to camera improvement: retaining the camera but subjecting it to a radical redesign, or removing the camera entirely—effectively turning the product into an AI-enabled headset worn on the face.

Regardless of the technical path chosen, one thing is clear: privacy is no longer merely a selling point for AI glasses, but the primary constraint defining their physical form.

Source: Agang Says