A new study of seven smart glasses brands reveals extensive app permissions, hidden trackers, and widespread reliance on cloud processing.

As AI-powered smart glasses edge closer to mainstream adoption, their ability to see, hear, and interpret the world around us is raising increasingly uncomfortable questions about privacy.

With Samsung reportedly preparing to launch its first Android XR smart glasses in November, new research from Cybernews suggests that existing manufacturers still have significant work to do when it comes to protecting users’ personal information.

The researchers examined seven smart glasses brands — Meta, Rokid, RayNeo, INMO, Solos, Even Realities, and Halliday — assessing everything from app permissions and tracking technologies to data retention policies and recording indicators.

The findings suggest that privacy risks extend well beyond the possibility of being recorded without consent.

The concerns come as the wearable technology market continues to evolve. As previously explored in TechTrends’ report on the VR and AR device market, manufacturers have long faced challenges in making immersive technology appealing to mainstream consumers. Privacy could become another significant barrier to adoption.

Smart glasses are collecting more than you might think

Among the seven brands examined, Meta stood out for requesting the greatest number of app permissions.

According to the Cybernews analysis, Meta’s companion app requested 70 permissions, including 18 classified as dangerous under Android’s permission system.

These included access to sensitive functions such as audio recording, SMS messages, and external storage.

Although requesting a permission doesn’t necessarily mean an app is misusing it, the sheer number of potential access points highlights the amount of personal information that could be available to smart glasses manufacturers.

Meanwhile, researchers identified six trackers each in the INMO Global and Solos AirGo companion apps.

Rokid was the only app in the study where no trackers were detected.

The cloud privacy problem

Perhaps the most significant finding concerns where AI processing actually happens.

All seven brands examined rely on cloud processing for their AI functionality.

None of the devices analyzed performed core AI tasks, such as voice processing, translation, or image analysis, entirely on the glasses themselves.

This reliance on external infrastructure raises questions about what information leaves the device, where it travels, and how long it remains accessible.

Cloud computing has long played an important role in making immersive technologies more powerful and accessible. As TechTrends previously explored in Scaling Up Immersive Tech With Blockchain, distributing processing workloads beyond individual devices can help overcome hardware limitations.

However, when those workloads involve potentially sensitive personal information, the benefits of cloud processing must be weighed against the associated privacy implications.

For consumers increasingly accustomed to AI assistants embedded in everyday devices, understanding these data flows could become an important consideration when choosing wearable technology.

When the recording light isn’t enough

Smart glasses have already sparked debate over the privacy of people who might be recorded without realizing it.

Recording indicator lights are intended to provide some reassurance, but the Cybernews research identified inconsistencies in how manufacturers handle them.

Only Meta and Rokid were confirmed to respond when their recording indicator lights were covered.

The researchers also found that five of the seven brands failed to specify data retention limits, leaving users with little clarity about how long their information might be stored.

Furthermore, only Meta, Rokid, and Even Realities provided privacy documentation specifically addressing their smart glasses.

Privacy needs to be part of the design

As smart glasses evolve from novelty gadgets into increasingly capable AI-powered wearables, privacy concerns are likely to become harder to ignore.

The technology promises considerable benefits, from real-time translation and hands-free assistance to more accessible ways of interacting with digital information.

Indeed, as explored in our recent article on turning biometric data into immersive VR experiences, the combination of wearable sensors and immersive technology is opening up new possibilities for personalized digital experiences.

But these capabilities also highlight how much increasingly intimate information wearable devices can potentially collect and process.

The challenges are not entirely new. As TechTrends previously reported in its coverage of Microsoft’s HoloLens technology, immersive headsets have long relied on sophisticated combinations of cameras, microphones, and environmental sensors to interpret their surroundings.

As these technologies become smaller, more discreet, and more widely adopted, questions about transparency and consent become increasingly important.

The Cybernews findings underline an important challenge for the industry: making smart glasses genuinely useful without turning them into another source of opaque personal data collection.

For manufacturers hoping to bring AI eyewear into everyday life, transparency around permissions, processing, and data retention may prove just as important as the technology itself.

The post Exposing the hidden privacy risks of AI smart glasses appeared first on Tech Trends.