Meta Platforms is weighing the possibility of substantial job reductions that could affect more than a fifth of its workforce as the company channels vast resources into artificial intelligence infrastructure and anticipates greater automation across its operations.
People familiar with internal discussions told Reuters that senior leaders have begun considering broad staffing cuts, although neither the precise scale nor the timing of any announcement has been finalised.
Executives have recently informed some managers that preparations should begin for potential reductions, signalling that the company is examining how to restructure teams in anticipation of productivity gains driven by AI-assisted work.
Responding to questions about the plans, Meta spokesperson Andy Stone rejected the characterisation of an imminent restructuring, stating:
“This is speculative reporting about theoretical approaches.”
If the cuts ultimately approach 20% of the workforce, they would represent the most significant downsizing at the company since a major restructuring in 2022 and 2023.
At the end of December Meta reported employing nearly 79,000 people. In November 2022 the company dismissed roughly 11,000 employees, about 13% of staff at the time, before announcing a further 10,000 job reductions around four months later during a period executives described as the “year of efficiency”.
AI ambitions reshape corporate priorities
The potential layoffs come as chief executive Mark Zuckerberg intensifies efforts to position Meta as a leading force in generative artificial intelligence.
Over the past year the company has sought to recruit leading researchers by offering compensation packages that in some cases reach hundreds of millions of dollars over a four-year period, particularly for specialists joining a new internal superintelligence team.
Meta has also outlined an enormous capital commitment to expand its computing infrastructure. The company has said it intends to invest as much as $600 billion in new data centres by 2028 to support its AI ambitions.
Alongside that spending, Meta has pursued acquisitions aimed at strengthening its technology ecosystem. Earlier in the week it purchased Moltbook, a social network designed for AI agents, while Reuters previously reported that the company is spending at least $2 billion to acquire the Chinese artificial intelligence startup Manus.
Zuckerberg has repeatedly suggested that the growing capabilities of AI systems could dramatically change the scale of human teams required to develop new products. In January he said he was beginning to see “projects that used to require big teams now be accomplished by a single very talented person”.
The company’s strategy mirrors a broader shift across the American technology sector, where executives increasingly cite rapid advances in AI as a reason to streamline workforces. In January Amazon confirmed plans to eliminate about 16,000 positions, roughly ten percent of its staff. Last month the financial technology firm Block reduced its workforce by nearly half, with its chief executive Jack Dorsey openly pointing to AI tools as a way for companies to achieve more with smaller teams.
Meta’s determination to invest heavily in artificial intelligence also follows setbacks related to its large language model programme. The company’s Llama 4 system faced criticism last year after researchers argued that early benchmarking results had been presented in a misleading way.
Meta later abandoned plans to release the most powerful version of the model, known as Behemoth, which had originally been scheduled for launch in the summer.
Engineers within the superintelligence group have since been working on a new model called Avocado in an attempt to restore the company’s technological momentum. However, people familiar with the project say the system has yet to achieve the level of performance executives had hoped for, adding further pressure as Meta seeks to redefine itself around artificial intelligence.
