Five years after Facebook deactivated its facial recognition system used for tagging people in photos on the social network, citing privacy and legal concerns and a desire to find the "right balance," the company is now planning to reintroduce the technology.

According to four individuals familiar with the confidential plans who were not authorized to speak publicly, Meta, Facebook's parent company, intends to add this feature to its smart glasses, produced in partnership with the group that owns Ray-Ban and Oakley, potentially as early as this year. Internally referred to as "Name Tag," the feature would allow wearers of the smart glasses to identify individuals and access related information using Meta's AI assistant.

Meta's plans remain subject to change. An internal document reviewed by The New York Times indicates that the Silicon Valley company has been deliberating since early last year on how to launch this feature, which carries "safety and privacy risks." The May document outlined a plan to initially offer the "Name Tag" feature to attendees at a conference for the blind—a step Meta did not take last year—before a broader public release.

An internal Meta memo suggested that political instability in the United States could provide favorable timing for the feature's launch.

Meta's Reality Labs division, responsible for hardware like smart glasses, stated in the document, "We will launch during a volatile political period when many civil society groups that would typically criticize us will have their resources diverted by other matters."

Facial recognition technology has long raised civil rights and privacy concerns, with potential uses ranging from government surveillance and suppression of dissent to corporate tracking of unaware consumers and misuse by individuals in public settings. Due to accuracy concerns, some U.S. cities and states have restricted or banned police use of the technology. Recently, Democratic lawmakers called for U.S. Immigration and Customs Enforcement to cease using facial recognition on American streets.

Nathan Freed Wessler of the American Civil Liberties Union stated, "Facial recognition technology on U.S. streets poses a uniquely severe threat to the de facto anonymity we all rely on. This technology is highly susceptible to abuse."

Meta had considered adding facial recognition to the first generation of Ray-Ban smart glasses in 2021 but abandoned the effort due to technical challenges and ethical concerns. The company has since revived these efforts, influenced by the Trump administration's closer ties with big tech companies and the unexpected strong sales of Meta's smart glasses.

EssilorLuxottica, Meta's partner in producing the glasses, reported this week that smart glasses sales surpassed 7 million units last year.

Three sources indicated that Meta CEO Mark Zuckerberg supports incorporating facial recognition to differentiate the product and enhance the utility of the AI assistant within the glasses.

According to two individuals, Meta is researching whom the technology could identify. Potential scopes include individuals already connected to the user on Meta platforms and people the user may not know personally but who have public profiles on Meta-owned sites like Instagram.

Two informed sources clarified that the feature would not become a universal facial recognition tool allowing users to query information about anyone they encounter.

Meta said in a statement, "We build products to help millions of people connect and enrich their lives. While we often hear interest in such features, and similar products exist in the market, we are still exploring various options. Should we decide to proceed, we would do so thoughtfully."

Last year, The Information reported that Meta had restarted its smart glasses facial recognition project.

Meta's smart glasses have already been used for identification purposes. In 2024, two Harvard students used Ray-Ban Meta glasses equipped with the commercial facial recognition tool PimEyes to identify strangers on the Boston subway, posting a viral video about their experiment. Meta emphasized at the time that a small white LED light on the right top corner of the frames "notifies others when a user is recording."

Currently, Meta smart glasses require the wearer to actively wake the AI assistant to ask questions, take photos, or record videos. Three sources said the company is also developing glasses, internally called "super sensing," which would continuously use the camera and sensors to record the user's daily life, similar to how AI notetaking tools summarize video meetings.

Facial recognition would be a core feature of these "super sensing" glasses, for instance, by reminding the wearer of pending tasks when a colleague is recognized. One source mentioned that Zuckerberg has questioned whether such glasses should keep an LED light constantly illuminated to indicate the "super sensing" mode or use an alternative notification method.

Meta has been developing facial recognition technology for over a decade. Three informed sources stated that Zuckerberg supports using AI and facial recognition from the company's Fundamental AI Research (FAIR) lab to assist blind and low-vision individuals, including through collaborations with external organizations like the accessibility technology company Be My Eyes.

Mike Buckley, CEO of Be My Eyes, said he has been in discussions with Meta for "about a year" regarding facial recognition glasses for people with low vision or blindness. "For this community, the technology is profoundly meaningful and powerful," he noted.

Mark Riccobono, President of the National Federation of the Blind, expressed support for the idea, although he stated he was unaware of specific plans to provide glasses to attendees at the organization's conference this July.

Meta has faced significant costs related to privacy issues. In recent years, the company paid $2 billion in settlements in Illinois and Texas for collecting facial data without consent—related to the very Facebook facial recognition system it shut down, which was designed to make tagging friends in photos easier. In 2019, Facebook paid a $5 billion settlement to the Federal Trade Commission (FTC) to resolve charges of user privacy violations, including those involving its facial recognition software.

As part of the FTC settlement, Meta agreed to conduct privacy risk reviews for each new or modified product. However, an internal post from January 2025 indicated that Meta had relaxed its privacy risk review process: the privacy team's influence over product launches was reduced, and new time limits were set for risk assessments.

Around the same time, employees responsible for risk reviews questioned whether the revised process still complied with the FTC settlement terms. According to a recording of an internal meeting obtained by The New York Times, Andy Millian, head of risk review at Reality Labs, told staff she believed these changes would "challenge" the boundaries of Meta's agreement with the FTC. "Mark wants to push the envelope a little," Millian said, referring to Zuckerberg.