…Smart glasses, rings, watches quietly harvest personal data, study warns
A new analysis has raised fresh concerns over the privacy implications of wearable technology, spotlighting the massive amount of user data being collected by companion apps that support smart glasses, rings and watches.
The study, conducted by Surfshark, examined the data collection practices of 24 popular wearable companion apps available on the Apple App Store and the findings are startling.
The study is limited to iOS apps only, providing focused insight into their data practices. The companion apps analyzed were selected based on top search results for popular smart glasses, rings, and watches. Data was collected from the Apple App Store on July 27, 2025, where apps are required to disclose up to 35 different data types across 16 categories.
Read also: Quick loans, big secrets: How Nigerians are paying with their privacy
From tracking sensitive health data to gathering location, contacts, and even biometric identifiers, many apps collect far more data than users might expect. “In fact, some apps gather up to 33 out of 35 possible data types listed by the App Store, a staggering 94 percent, highlighting an urgent need for greater transparency and stronger data protection frameworks,” the study revealed.
Smart glasses topped the list as the most invasive category in terms of data collection. The analysis reviewed 10 companion apps for smart glasses, including Meta AI, Amazon Alexa, Solos AirGo, and VITURE’s SpaceWalker. On average, these apps collect nine different data types, but some go far beyond that.
The study cited Meta AI app, which powers wearables like the Ray-Ban Meta Smart Glasses, collecting a jaw-dropping 33 of 35 data types, including location, browsing history, contacts, financial info, and more.
Similarly, Amazon Alexa, used with Echo Frames, collects 29 data types.
However, not all apps were found to be data-hungry. Notably, SpaceWalker by VITURE and Rokid stated they do not collect any data, offering a rare privacy-conscious alternative in a space dominated by aggressive data practices.
Among the most worrying findings is the use of personal data for vaguely defined categories. The App Store labels some data uses as “Other Purposes,” with no specific explanation provided. Meta AI and Amazon Alexa were the only apps that admitted to using user data for this catch-all category, involving 25 and 28 unique data types respectively.
More concerning, Amazon Alexa’s data handling under other purposes may include sensitive information such as health status, racial background, religious beliefs, sexual orientation, and even genetic or biometric data.
The analysis also examined seven smart ring apps, including Ultrahuman, Oura, and Zepp. While most collect an average of six data types, Ultrahuman was the only app found to explicitly admit using data for third-party advertising, including identifiers like email addresses and device IDs.
Meanwhile, all eight smartwatches analyzed, including Fitbit, Garmin Connect, and Huawei Health, collect user data.
Fitbit: Health & Fitness, owned by Google, emerged as the most data-intensive app in this category, potentially gathering up to 21 data types, nearly double the category average. Others, like CMF Watch and CASIO WATCHES, also collect data that may be used for tracking purposes.
The findings raise broader questions about whether wearables, now embedded in everyday life and used for everything from health monitoring to productivity could become silent surveillance devices.
Read also: The future of AI-based travel: A threat to privacy or a necessity?
“Because many consumers wear multiple devices simultaneously, such as a smart ring, smart glasses, and a smartwatch, the risk of privacy erosion compounds quickly,” the report noted.
The growing use of AI integration in wearables adds another layer of complexity, potentially exposing users’ unchangeable biometric data.
Making matters worse, the study highlights the growing problem of privacy fatigue, the exhaustion users feel from managing long and complex privacy policies.
According to a separate research article cited in the study, the longest policy analyzed was 12,125 words, and even the shortest spanned over 4,400 words, making it difficult for consumers to make informed choices.
