Meta just told investors it will spend up to $145 billion on AI this year, and markets punished the company for it. Even as Meta AI spending projections for 2026 showed revenue up 33%, its fastest quarterly growth since 2021, shares still fell more than 7%. Wall Street’s message was blunt: the numbers don’t justify the spend.
CEO Mark Zuckerberg blamed the spike on “higher component costs, particularly memory pricing.” That phrase reaches far beyond Silicon Valley. The AI boom has triggered an unprecedented data centre buildout, squeezing global memory chip supply and raising prices worldwide. Nigerians buying laptops or smartphones are already feeling this pressure. Global memory chip shortages have driven up device prices worldwide, and Nigeria, which imports most of its consumer electronics, absorbs those increases directly.
Why Meta Is Doubling Down and Why It’s Behind
Meta is projecting to spend nearly twice what it recorded in 2025. The acceleration looks less like a strategy and more like a company that knows it’s falling behind. Google has surged ahead in the AI race, and Zuckerberg spent most of Wednesday’s earnings call trying to reassure investors that his company has a credible plan.
The catch-up effort was launched about ten months ago. The effort included aggressive talent poaching. Zuckerberg brought in Scale AI founder Alexandr Wang to lead the new Meta Superintelligence Labs division.
That division produced its first public output this month, an AI model called Muse Spark, which Meta plans to open-source eventually. Zuckerberg told investors: “This was the first release from Meta Superintelligence Labs, and it shows that our work is on track to build a leading lab.” One model release doesn’t close the gap on Google.
See Also: Nigeria Orders Banks and Fintechs to Disclose Cyberattacks
AI Agents, Workforce Cuts, and What Comes Next
Meta’s next move involves two AI agents, one built for personal use and one for businesses. The business agent is already in testing. Zuckerberg said weekly conversations with that product have grown tenfold since the start of 2026.
Each week, over 500 million users across Facebook and Instagram consume AI-translated and dubbed video content. Meta’s recommendation systems are getting an AI overhaul. The goal: hyper-personalize every user’s feed.
Internally, the transformation is sharper still. Meta has announced a 10% headcount reduction, with voluntary exit packages offered to 7% of its American employees. Li declined to link the cuts to automation directly. She did say a “leaner operating model” would help offset the company’s massive capital outlays, which is as close to a confirmation as corporate language gets.
See Also: Kenyan AI Startup Lua Raises $5.8M to Replace Human Workflows
The Metaverse Hangover Is Still Bleeding
Any fair assessment of Meta’s AI spending has to contend with what came before it. The Reality Labs division, the vehicle for Zuckerberg’s metaverse vision, posted an operating loss of more than $4 billion this quarter on just $402 million in revenue. Over six years, that division has lost more than $80 billion in total.
That history is why investors are nervous. The metaverse promised a new computing paradigm and delivered VR headsets that most people never used. AI is a different bet, more embedded in existing products, more measurable, and more aligned with where users already spend time.
But $145 billion demands results, and Zuckerberg is working against a credibility deficit he built himself. By year-end, the results or lack of them will be impossible to hide. Meta is spending like it has no choice because, arguably, it does not.
Want sharper takes on the tech decisions shaping Africa and the world? Join the GizPulse newsletter: one email, twice a week, no noise.
