Meta’s Embrace of A.I. Is Making Its Employees Miserable, Here’s What’s Really Happening

“This makes me super uncomfortable.” That was the reaction of one Meta engineering manager when the company announced it would start recording every mouse movement, keystroke, and screen click on employee laptops. No opt‑out. No negotiation. Just a memo saying your daily digital life was now training data for the very AI systems that might one day replace you. Welcome to Meta in 2026, a company so hell‑bent on winning the AI race that it has turned its own workforce into exhausted, anxious, and increasingly resentful participants in an experiment they never signed up for.

The “Year of Intensity” That Never Ended

Cast your mind back to January 2025. Mark Zuckerberg warned employees to “buckle up” for a year of high intensity. What followed was a relentless cycle of reorganisations, pivots, and pink slips. By mid‑2025, Meta had poured $14.3 billion into Scale AI, absorbed its founder Alexandr Wang as Chief AI Officer, and created the Meta Superintelligence Labs (MSL). The mission? Build artificial general intelligence before anyone else.

But the cost was staggering. Reality Labs haemorrhaged $4.4 billion in a single quarter while generating just $470 million in revenue. To offset the bleeding, Meta slashed 600 AI roles in October 2025, then another 10 % of its entire workforce, roughly 8,000 people, in May 2026, while closing 6,000 open positions. That “year of intensity” never ended; it just became the new normal.

When Your Mouse Clicks Become Training Data

The Tool That Feels “Very Dystopian”

In April 2026, Meta rolled out the Model Capability Initiative (MCI), software that captures exactly how employees move their mouse, click buttons, navigate dropdown menus, and even what appears on their screens. The company’s pitch was almost poetic: “This is where all Meta employees can help our models get better simply by doing their daily work.”

But employees saw it differently. One worker told the BBC the programme felt “very dystopian.” Anonymous posts on Blind exploded with anger. The feeling of being reduced to a data‑generating machine, knowing that every action you take is teaching an AI to mimic, and eventually replace, you, is a special kind of workplace horror. And when CTO Andrew Bosworth confirmed there was “no option to opt‑out,” the message was clear: your body of work is now a free training set, and your consent is irrelevant.

AI as the New Performance Scorecard

If the tracking tool felt invasive, the performance overhaul felt like a trap. In November 2025, Meta’s Head of People Janelle Gale told employees that “technology‑driven impact” would become a formal evaluation criterion in 2026. Translation: your promotion, bonus, and even your job security would soon depend on how enthusiastically you used AI.

Initially, the company encouraged workers to sprinkle “AI‑fuelled wins” into their self‑evaluations. Then came Metamate, an internal AI chatbot designed to draft performance reviews by scraping an employee’s project history and peer feedback. The irony was almost too perfect: an AI system would now summarise your work for a manager who might also be using AI to evaluate it. By late 2025, explicit AI‑usage targets were being set, some engineers were told 50, 80 % of their code must be AI‑assisted. One leaked document even specified that 55 % of code modifications must involve an AI agent.

Employees described the pressure as “suffocating.” The message from leadership felt unambiguous: use AI or get out.

The Talent War That Set Employees Against Each Other

Meanwhile, a parallel drama was unfolding at the top of the food chain. To staff its Superintelligence Labs, Meta went on a hiring spree that saw it poach star researchers from OpenAI, Google, and Apple with signing bonuses reportedly reaching $100 million and multi‑year packages of up to $300 million.

For existing Meta AI researchers, this was devastating. “Bringing in outsiders for 10 to 50 times higher compensation made it feel like Zuckerberg was telling GenAI employees they had failed,” a former Meta AI researcher told Business Insider. The resentment was not just about money, it was about prestige, compute resources, and the sense of being demoted in your own company. Some employees threatened to quit just to negotiate a spot on the elite MSL team; others left outright.

The psychological damage went deep. A former Meta research scientist published a scathing essay describing a “culture of fear” that was “spreading like cancer.” Glassdoor reviews painted a similar picture: “Great products, terrible culture… employees are encouraged to compete instead of collaborating.” Even the anonymous forum Blind showed a 300 % jump in negative sentiment about AI at Meta between 2024 and late 2025, with 83 % of AI‑related posts now carrying negative tone.

The Productivity Paradox, More AI, More Burnout

It’s worth stepping back and asking: does any of this even work? The uncomfortable truth is that AI’s productivity promise comes with a dark side. Researchers from Harvard Business Review coined the term “brain overload” to describe what happens when workers constantly switch between multiple AI tools, decision fatigue increases, errors multiply, and mental exhaustion sets in. About 14 % of workers in one large study reported experiencing “mental fatigue that results from excessive use of, interaction with, and/or oversight of AI tools.”

Columbia Business School professors have found that AI‑related stress is fundamentally different from past waves of technological change. “Change is hard, but the really hard part is if you don’t know which direction things are going to change,” explains Professor Stephan Meier. And what makes AI uniquely anxiety‑inducing is that it attacks the very core of professional identity, the skills and expertise that workers spent decades building.

At Meta, this paradox is playing out in real time. Employees are expected to produce more with AI, yet the cognitive load of babysitting AI outputs, fixing hallucinations, and meeting arbitrary usage quotas is leaving them more drained than ever. One worker on Hacker News captured it perfectly: “Since adopting AI at work, expectations have tripled, stress has tripled, but actual productivity is up only about 10 %.”

What This Means for Tech Workers Everywhere

Meta’s story is dramatic, but it’s not unique. Microsoft, Amazon, and Dell are all running their own versions of the AI‑efficiency playbook. The warning signs are universal: when companies treat employees as raw material for AI training, morale collapses. When performance metrics prioritise AI usage over human judgment, creativity suffers. And when the threat of layoffs hangs over every internal reorganisation, trust evaporates.

If you’re a tech worker feeling uneasy about your company’s AI push, you are not alone, and you are not irrational. The research backs up what your gut is telling you: the way these transitions are being managed is hurting people.

The question that remains is whether companies like Meta will recognise the damage before it’s too late. A 43 % drop in culture rating since 2020 is not a blip, it’s a screaming siren. Whether leadership chooses to hear it is the next chapter.