On March 14, Reuters reported that Meta is gearing up for another significant round of layoffs, which could impact up to 20% or even more of its workforce.

As of December 31, 2025, Meta employed 78,865 people. This means that nearly 16,000 individuals may soon find themselves out of work. Sources indicate that Meta's top executives have already briefed senior management on the layoff plan and have asked them to come up with proposals for streamlining teams.

This will mark Meta's largest round of layoffs since its restructuring at the end of 2022, dubbed the "Year of Efficiency," and will far exceed the combined total of the previous two rounds. What's particularly striking is that these layoffs are happening not during a downturn, but against the backdrop of Meta's record-breaking performance—its full-year revenue for 2025 surpassed USD 200 billion for the first time, reaching USD 200.966 billion.

On one hand, there's record-breaking revenue; on the other, the largest layoffs in the company's history. This begs the question: what's really going on inside Meta?

Back in January of this year, Meta CEO Mark Zuckerberg made a statement in an internal meeting that now seems eerily prescient: "We're starting to see that projects that once required multiple teams to collaborate on can now be completed by a single highly capable individual."

At the time, this was seen as an optimistic view of AI-driven efficiency. But now, it serves as the perfect footnote to the large-scale layoffs. Zuckerberg wasn't just speaking theoretically; he was describing a reality that's already unfolding—AI tools are reshaping how companies view human efficiency.

Indeed, since the start of this year, the wave of layoffs in the U.S. tech sector has picked up steam, with more and more companies directly citing AI as a reason for the cuts. Amazon confirmed in January that it would lay off around 16,000 employees, while fintech company Block laid off nearly half of its workforce last month. Block's CEO, Jack Dorsey, explicitly stated that the growing power of AI tools allows the company to achieve more with fewer people.

As AI evolves from a supplementary tool to the core of productivity, the survival space for traditional roles is shrinking rapidly.

It's reported that the direct catalyst for Meta's large-scale layoffs is its massive investment in artificial intelligence infrastructure. The company plans to invest over USD 600 billion in the United States by 2028 to build AI data centers and computing infrastructure.

This sum is three times Meta's full-year revenue for 2025 and more than eight times its 2025 capital expenditures (USD 72.2 billion). In 2026 alone, Meta's capital expenditures are expected to reach between USD 115 billion and USD 135 billion, nearly double those of 2025.

These funds will be primarily used for constructing AI data centers, computing facilities, procuring third-party cloud services, building AI model training infrastructure, and recruiting top AI talent.

To maintain a leading edge in this AI arms race, Meta is sparing no expense in luring top talent away from competitors. According to The New York Times, Meta has offered four-year contracts worth between USD 100 million and USD 300 million to top researchers poached from OpenAI or Google DeepMind. This "fewer people, more money" talent strategy stands in stark contrast to the large-scale layoffs of ordinary positions.

On one hand, there are multimillion-dollar compensation packages; on the other, layoff notices for tens of thousands. Meta's workforce structure is undergoing a ruthless restructuring.

During the earnings call in January of this year, Zuckerberg said, "We delivered strong business performance in 2025, and I look forward to driving the development of global personal superintelligence in 2026." Reading this statement now takes on a whole new meaning—the cost of driving the development of superintelligence may be the careers of tens of thousands of employees.

This AI-driven efficiency revolution is reshaping the world of work at a pace beyond our wildest imagination. As machines become increasingly intelligent, how should we redefine human value? This is not just a question for Meta to answer; it's a challenge each of us must confront.