Meta’s New Privacy Direction: Less Control, More Data
Meta’s latest privacy changes combine the removal of its Off-Facebook Activity control with a quietly deployed facial recognition system in its smart glasses app, signaling a strategic shift toward deeper, more opaque data collection across both its social platforms and emerging wearable devices. Instead of building new ways for people to manage how they are tracked, Meta is folding existing protections into vague personalization settings and experimenting with face-based surveillance baked into popular hardware. This is not a minor interface tweak; it is a philosophical turn away from the earlier promise that users could meaningfully limit Meta privacy tracking. The practical result is simple: your ability to keep Meta from stitching together your movements on the web and in the physical world is shrinking, even as its cameras and sensors spread.
Off-Facebook Activity Removal: From Real Opt-Out to Cosmetic Toggle
When Meta launched the Off-Facebook Activity control, it sold it as putting “people’s control over their data” ahead of profits, allowing users to disconnect identifying information from data that apps and websites sent to Facebook. That feature is now “going away,” replaced in July by an expanded “Activity from other businesses” setting that only lets you choose whether Meta uses this incoming data to show you ads and other content, not whether it gets the data in the first place. Meta’s own note concedes that “regardless of your choice, businesses and organizations may send us your activity, and we may anonymize it or use it to improve our products.” In other words, the old Off-Facebook Activity removal actually cut a link between your account and external tracking; the new control is a veneer on top of continued data collection control by Meta, with personalization as the default justification.
This streamlining is framed as less confusing for users, but the substance matters: you no longer have a built-in, one-click way to disconnect off-Meta browsing from your identity. Instead, your choices narrow to how your off-site behavior is used to shape what you see, while Meta privacy tracking via pixels, buttons, and embeds keeps flowing in the background. The company insists it is not gathering new categories of data, yet the decision to retire a privacy tool that once sat proudly on its integrity timeline shows where priorities now lie. Privacy is being converted from a structural safeguard into a UX slider.
Facial Recognition in Smart Glasses: NameTag and a Wearable Panopticon
Alongside the tracking rollback, Meta’s expanding smart glasses ecosystem is testing something even more intrusive: face-based identification in public. Meta’s Ray-Ban glasses sold more than seven million units in 2025, giving the company roughly 82 percent of the smart glasses market and a rapidly growing camera fleet on people’s faces. In June, researchers at a digital rights group found a dormant system called NameTag hidden in an update to the Meta AI glasses app on over 50 million phones. NameTag uses three AI models in sequence—detecting faces, turning each into a 2,048-number faceprint, and matching against stored databases—making real-time facial recognition smart glasses a practical product feature, not science fiction. When a technologist enabled it in debug mode, the app identified a known face on the fly and was described as “nearly ready to go,” a step toward turning wearers into what he called a “distributed surveillance network.”
Meta stripped the NameTag code a day after the findings were published, without a public explanation. That is not reassuring; it only proves the company is willing to ship facial recognition systems at scale, then quietly retreat when noticed. Prior work has already shown how dangerous this can be. In 2024, two students used Meta Ray-Ban glasses plus a custom reverse facial recognition engine to identify strangers in under a minute, pulling their names, home addresses, occupations, and partial Social Security numbers from public databases. The students demonstrated greeting subway riders by name to highlight the threat, not to commercialize it—but Meta now has similar building blocks sitting inside its own software. The privacy question is no longer whether this can be done but whether, and when, Meta will normalize it as a product feature.
What Ordinary Users Can Still Do—and Why It Isn’t Enough
Even as Meta reshapes its controls, you are not completely powerless. You can still toggle the updated “Activity from other businesses” setting so Meta does not use your off-site activity to personalize ads and content, pushing the platform to rely more on what you do on its own services—likes, reels, posts—for targeting. You can also blunt Meta privacy tracking by using privacy-optimized browsers instead of mainstream options that are friendly to third-party cookies and trackers. One digital rights engineer recommends tools such as a privacy-focused browser extension that blocks Meta embeds from loading until you explicitly activate them, and system-level phone settings that restrict ad tracking across apps. These measures reduce the data stream flowing into Meta’s systems, and they are worth the effort.
Still, the direction of travel is clear: Meta is recentering control on its own internal toggles while leaving underlying collection intact. Wearables add a new dimension, where you can be recorded or potentially face-scanned by someone else’s device even if you never agreed to the platform’s terms. No browser setting stops a stranger’s glasses from filming you in a shopping mall. The combination of Off-Facebook Activity removal and experiments with facial recognition smart glasses shows a platform shifting its balance from user protection to product expansion. Individual defenses can slow that shift but cannot fully counteract it while the core business model prizes behavioral data above consent.
The Bigger Picture: Meta’s Bet Against Privacy
Taken together, these moves signal that Meta is betting users will tolerate less meaningful data collection control in exchange for convenience, personalization, and shiny devices. Off-Facebook Activity began as a rare admission that privacy can hurt revenue and still be worth doing; now it is being folded into a softer preference menu that keeps the data flowing. NameTag’s appearance in an app update already on more than 50 million phones shows the company is willing to ship advanced surveillance capabilities first and explain later. That sequence—deploy, get caught, silently pull back—should worry anyone who cares about how face data might be used when the next update lands and news attention has drifted.
The conclusion is uncomfortable but straightforward: Meta is dismantling its own privacy guardrails as it builds a world where its cameras and trackers follow you online and off, even through other people’s devices. Users can and should respond by tightening browser settings, installing privacy extensions, and resisting facial recognition smart glasses in their social circles. But lasting protection will require more than individual action; it will mean demanding transparent limits on what data Meta can collect, how long it can keep it, and whether opt-out can still mean “stop gathering this,” not just “use it differently.” Until then, every new product launch and settings change should be read with skepticism: not “how does this help me?” but “what new thing does this let Meta learn about me?”
