Meta Cuts 700 Jobs as Spending Shifts to $135 Billion AI Infrastructure Buildout

0
6

Now the “Year of Efficiency” has evolved into a permanent strategy of lean operations for the social media giant. On Wednesday, March 25, 2026, Meta Platforms began laying off approximately 700 employees across several key divisions, including Reality Labs, recruiting, and sales. Therefore, this restructuring marks a decisive pivot away from money-losing metaverse projects toward a massive $135 billion capital expenditure plan for artificial intelligence. Currently, as the company prepares for the mid-2026 launch of its flagship “Avocado” AI model, leadership is aggressively reallocating payroll costs into high-performance compute capacity. Thus, while hundreds face role eliminations, the company is simultaneously hiring top-tier AI talent to maintain its edge against OpenAI and Google.

Also Read | Imran Khan and Bushra Bibi Sentenced to 17 Years in Jail

- Advertisement -

 

At a Glance:

  • The Layoffs: Approximately 700 roles impacted, primarily in Reality Labs and recruiting.

  • The AI Bet: Meta projects spending between $115B and $135B on AI infrastructure in 2026.

  • Infrastructure News: A massive $27 billion deal was recently signed with Nebius for Nvidia “Vera Rubin” GPU capacity.

  • The Metaverse Scale-Back: Reality Labs, which has lost over $80 billion since 2021, remains the hardest-hit division.

  • Workforce Status: Meta started 2026 with roughly 79,000 employees, despite ongoing “flattening” of teams.

In This Article:

  • The 700-Person Cut: Which Teams Are Most Affected?

  • From Labor to Compute: Funding the $135 Billion AI Goal

  • The $27 Billion Nebius Deal: Securing the Future of GPUs

  • Reality Labs Restructuring: Is the Metaverse Vision Fading?

  • Frequently Asked Questions (FAQs)

The 700-Person Cut: Which Teams Are Most Affected?

Now the latest workforce reduction is being described as a “streamlining” effort to align with the company’s 2026 goals. Because Mark Zuckerberg has committed to “flattening” the organization, redundant managerial and support roles are being phased out.

First, the Reality Labs division saw its second major cut of the year, following 1,000 layoffs in January. Next, recruiting and sales teams were significantly impacted as the company automates internal workflows using AI agents. Thus, the cuts are not a sign of financial distress but rather a structural shift toward a more technical, AI-centric workforce. Currently, a Meta spokesperson confirmed that the company is attempting to find “other opportunities” within the organization for those impacted where possible.

Also Read | Imran Khan and Bushra Bibi Sentenced to 17 Years in Jail

 

From Labor to Compute: Funding the $135 Billion AI Goal

Now the financial logic behind these layoffs is explicit: Meta is converting labor costs into raw computing power. Because the “AI arms race” requires unprecedented capital, the company is raising its 2026 expenditure forecast to nearly double 2025 levels.

First, Meta plans to spend up to $135 billion this year on data centers and advanced chips. Next, CFO Susan Li noted that these investments are critical for training the upcoming Llama 4 models and supporting “Personal Superintelligence” features. Thus, every dollar saved on non-core roles is being redirected into the hardware needed to outpace competitors like Anthropic. So, investors have reacted positively to this discipline, with Meta’s stock climbing 3% following the layoff reports.

 

The $27 Billion Nebius Deal: Securing the Future of GPUs

Now the scale of Meta’s infrastructure ambition was further revealed by a massive partnership agreement. Because high-performance GPUs are becoming a scarce global commodity, Meta has moved to secure its supply chain.

First, the company signed a $27 billion capacity deal with neocloud provider Nebius Group. Next, this agreement grants Meta immediate access to dedicated clusters powered by Nvidia’s next-generation Vera Rubin architecture. Thus, by partnering with specialized providers, Meta avoids the time-consuming process of building every data center internally. Currently, this deal is one of the largest private sector commitments to AI infrastructure ever recorded, ensuring Meta remains a “hyperscaler” in the generative AI era.

Also Read | Imran Khan and Bushra Bibi Sentenced to 17 Years in Jail

Reality Labs Restructuring: Is the Metaverse Vision Fading?

Now the division once central to Meta’s identity is facing a harsh reality. Because Reality Labs has incurred cumulative losses exceeding $80 billion, the company is narrowing its focus within the unit.

First, Meta has reportedly shut down three virtual reality studios and paused work on its workplace-oriented metaverse platforms. Next, the focus is shifting from immersive VR worlds toward AI-powered wearables, such as smart glasses with cameras and real-time translation. Thus, while the “Metaverse” name remains, the product roadmap is being reshaped to fit an AI-first world. Currently, the goal is to integrate AI assistants directly into hardware, potentially turning Reality Labs from a loss-leader into a profitable hardware division.

 

Frequently Asked Questions (FAQs)

How many people were laid off in the latest Meta round?

Approximately 700 employees were impacted in this round, following a larger cut of 1,000 roles in January 2026.

Why is Meta laying off staff if they are making record profits?

The company is redirecting funds from traditional departments like recruiting and social media sales into AI infrastructure and talent to stay competitive in the AI race.

What is the “Avocado” model mentioned in reports?

“Avocado” is the rumored internal codename for Meta’s next-generation flagship AI model, expected to launch in mid-2026 to compete with Google Gemini and GPT-5.

Is Meta abandoning the Metaverse?

Not entirely, but it is scaling back significantly. The company is shifting focus from VR gaming and virtual worlds to AI-integrated hardware like smart glasses.

What is the $27 billion deal with Nebius?

It is a long-term agreement where Meta pays for dedicated access to Nvidia-powered GPU clusters, ensuring they have the computing power needed to train and run massive AI models.

Also Read | Imran Khan and Bushra Bibi Sentenced to 17 Years in Jail

End…

- Advertisement -