Meta’s A.I. Push Faces Internal Turmoil and Privacy Fears

Meta’s AI Drive Runs Into Internal Upheaval and New Privacy Questions

Meta, already spending heavily to recast itself around artificial intelligence, is facing a fresh wave of pressure over how that strategy is unfolding — both inside the company and in the products it is building.

In recent days, reports have described a company straining under the weight of a sweeping A.I. reorganization, with employees complaining of confusion, forced job moves and deepening mistrust of leadership. At the same time, new details about Meta’s work on facial recognition for its smart glasses have revived concerns that the company is moving closer to surveillance-sensitive technology than it has publicly suggested.

Together, the developments point to a more unsettled picture of Meta’s A.I. push: ambitious, expensive and increasingly controversial.

A Reorganization That Has Shaken Morale

The internal friction follows Meta’s broad May restructuring, which put artificial intelligence at the center of the company’s operations. The overhaul included layoffs affecting about 10 percent of staff and the reassignment of roughly 7,000 employees into A.I.-related efforts, according to prior reporting. The company has also signaled the scale of its commitment in financial terms, raising its 2026 capital-expenditure forecast to between $125 billion and $145 billion, much of it tied to A.I. infrastructure.

But the transition has been rocky.

Employees in Meta’s new Applied AI organization have described a workplace marked by volatile management decisions, unclear reporting lines and resentment over involuntary transfers, according to internal discussions and accounts reviewed by WIRED. The tensions have spilled into meetings and internal forums, where frustration with executives has become unusually blunt.

The unrest appears to reflect a broader problem confronting large technology companies in the A.I. race: executives want speed, focus and reallocation of talent, while workers who have already endured layoffs and repeated strategic pivots are being asked to do more with less certainty.

At Meta, that tension has become especially visible because the company is trying to transform itself on several fronts at once — from social networking giant to A.I. platform, hardware maker and builder of so-called A.I. agents. The result, critics inside the company say, is an organization being rapidly rewired without enough stability to make the new structure function smoothly.

A Hackathon Meets Employee Cynicism

That strain has also colored the reception to Mark Zuckerberg’s plan for a companywide A.I. hackathon scheduled for mid-July.

Hackathons have long been part of Silicon Valley mythology, framed as exercises in creativity and urgency. But inside Meta, some employees reportedly saw the event less as inspiration than as a symbol of how detached management had become from daily realities after layoffs, heavier workloads and low morale.

One employee, writing in a companywide forum, questioned whether Meta still had the kind of culture that could support such an event. The criticism underscored a larger complaint: that leadership is trying to impose the rituals of startup intensity on a work force that has been through repeated rounds of disruption.

Zuckerberg has reportedly promised changes intended to steady the company, including more role stability, lower manager-to-staff ratios and greater team spending. Whether those steps will calm the discontent remains unclear. For many employees, the hackathon may become an early test of whether Meta’s leadership can rally the company behind its A.I. agenda — or whether the event will simply expose the widening gap between executive ambition and employee sentiment.

Smart Glasses and a Familiar Privacy Flashpoint

Outside the company, Meta is confronting another source of scrutiny: facial recognition.

The company licensed facial-recognition software from Rank One Computing, a supplier with ties to defense and law-enforcement work, for internal development related to its smart-glasses app, according to a report published Sunday. Rank One’s board includes former national security and law-enforcement officials, a detail likely to sharpen questions about the nature of the work and the kinds of applications Meta may have been exploring.

The disclosure adds a more concrete dimension to a debate that has been building for months around Meta’s smart-glasses ambitions. Earlier reporting this year said the company had been considering a feature, known internally as “NameTag,” that could identify people through the glasses. On June 4, WIRED reported that dormant face-recognition components had already been embedded in the Meta AI companion app. Days later, after questions were raised publicly, the company removed that code.

What remains uncertain is whether Meta has abandoned the effort, paused it or simply moved it out of view. The company reportedly did not explain why the code had been removed, whether its relationship with Rank One was continuing or whether the feature could eventually return.

That ambiguity matters. Meta has a long and troubled history with facial recognition, including years of criticism and legal challenges over how it handled biometric data on Facebook. The company shut down its facial-recognition system on Facebook in 2021, presenting the move at the time as a response to growing public unease and regulatory uncertainty. Any renewed interest in the technology — especially in wearable devices capable of identifying people in real time — is likely to invite intense scrutiny from privacy advocates and lawmakers.

Why the Moment Is Especially Fraught

The timing is significant. Across the technology industry, companies are racing to embed A.I. into consumer products, workplace software and new hardware. Meta has tried to position itself as one of the few companies with the scale to compete on all three. It is investing heavily in chips, data centers, A.I. assistants and smart glasses, betting that the next era of computing will be shaped by generative A.I. and always-on wearable devices.

But those bets carry risks that are not merely technical.

Inside Meta, the challenge is execution: whether a company this large can rapidly reorganize around A.I. without damaging morale, productivity and trust. Outside Meta, the challenge is legitimacy: whether consumers and regulators will accept increasingly intimate A.I. products from a company whose record on privacy has often fueled skepticism.

The latest reports suggest those two pressures are beginning to converge. A company trying to move faster on A.I. is doing so amid internal disorder, while one of its most sensitive product areas — smart glasses equipped with potentially identifying capabilities — is reviving old fears about surveillance.

For Meta, the question is no longer just whether it can lead in artificial intelligence. It is whether it can do so without deepening the operational and political problems that have followed the company for years.

Sources

Further reading and reporting used to add context:

AI News