Meta is cutting 10% of its global workforce, roughly 8,000 employees, while simultaneously committing between $115 billion and $135 billion to artificial intelligence infrastructure this year alone. The announcement, made by Meta's chief people officer Janelle Gale in an internal memo, represents one of the most aggressive bets of Mark Zuckerberg's tenure as CEO. The layoffs take effect May 20, and the company is also closing 6,000 open positions it had planned to fill.

Why Is Meta Cutting Jobs While Reporting Record Profits?

The timing might seem counterintuitive. Meta reported record revenue of nearly $201 billion in 2025 and record fourth-quarter net income of $22.8 billion. Yet the company is shedding headcount at an accelerating pace. The answer lies in where the money is going. Meta's capital expenditure is nearly doubling from $72.2 billion in 2025 to between $115 billion and $135 billion in 2026, with almost all of it directed toward artificial intelligence data centers, graphics processing units (GPUs), and the infrastructure powering its Llama models.

This is not Meta's first round of cuts in 2026. Earlier in the year, the company eliminated roughly 1,000 to 1,500 positions inside Reality Labs, the unit responsible for virtual reality and metaverse efforts, shutting down several VR game studios in the process. A second, smaller round followed in March, reaching hundreds of employees across Facebook, Reality Labs, global operations, and sales.

What Is Meta's Strategy for Replacing Human Workers With AI?

Meta's restructuring reflects a deliberate shift toward automating tasks that humans currently perform. The company last year acquired a 49% stake in data-labeling firm Scale AI for more than $14 billion, bringing Scale's former CEO Alexandr Wang in-house to lead Meta Superintelligence Labs. The company has also moved to replace third-party content moderation contractors with artificial intelligence-powered systems, further reducing its dependence on human labor at scale.

Days before the layoff announcement, Meta revealed an internal monitoring program called the Model Capability Initiative (MCI) that has already drawn sharp internal backlash. The program installs tracking software on U.S.-based employees' work computers, logging mouse movements, click locations, keystrokes across hundreds of approved apps and websites, and periodic screenshots capturing on-screen activity in context.

How Meta Is Training AI to Replicate Human Computer Use

• Keystroke Tracking: The Model Capability Initiative logs keystrokes across Google, LinkedIn, GitHub, Slack, Wikipedia, and Meta's own internal tools to understand how humans navigate software.
• Mouse Movement Monitoring: The system records mouse movements and click locations to teach AI agents how users interact with dropdown menus and interface elements.
• Screenshot Capture: Periodic screenshots capture on-screen activity in context, allowing AI systems to learn the visual patterns of human computer work.

Meta says the data will not be used for performance evaluations and that safeguards protect sensitive content. However, the timing, just days before a company-wide layoff, has deepened unease inside the company. Multiple employees described the program in internal messages as dystopian, raising concerns about exposure of sensitive personal data.

The stated purpose is to teach artificial intelligence agents how humans actually use computers, so those agents can eventually replicate the same tasks independently. This represents a fundamental shift in how Meta views its workforce: not as irreplaceable talent, but as training data for the systems that will eventually replace them.

Is This Pattern Unique to Meta?

Meta is not alone in this pattern. The tech industry has shed more than 95,000 jobs across hundreds of layoff events so far in 2026. Amazon announced the elimination of roughly 16,000 corporate positions in January. Microsoft this week offered voluntary buyouts to approximately 7% of eligible U.S. employees, a first for the 51-year-old company. Oracle cut up to 30,000 positions to fund its own artificial intelligence infrastructure push.

The pattern is consistent across all of them: record revenues, rising profits, and fewer people. This suggests a fundamental restructuring of the technology industry, where artificial intelligence infrastructure investment is becoming the primary driver of capital allocation, and human headcount is being treated as a cost to be minimized rather than an asset to be grown.

Meta is scheduled to report first-quarter 2026 earnings on Wednesday, April 29, alongside Alphabet, Amazon, and Microsoft. The company's shares fell 2.4% on Thursday and remain roughly flat for the year.