Meta has said it could spend between $60 billion and $135 billion in 2026 to build more powerful AI systems. This massive amount of money will go toward buying computer chips (GPUs), building huge data centers, and creating stronger AI tools. During their latest earnings call, CEO Mark Zuckerberg and CFO Susan Li explained that this big spending increase is needed to stay ahead in the fast-growing world of artificial intelligence.
In 2025, Meta already spent around $60–65 billion on similar things – one of the highest amounts among big tech companies. The 2026 plan could almost double (or even more than double) that amount.
Why Meta is spending so much
Meta shared these main reasons for the huge investment:
– Training and improving their open-source AI models (like the upcoming Llama 4)
– Building what could become one of the world’s biggest private AI supercomputers (with hundreds of thousands or even millions of GPUs)
– Adding more AI features to Instagram, WhatsApp, Facebook, Messenger, and their Meta AI assistant (things like better image/video creation, real-time answers, and tools for creators)
– Owning their own data centers and power supply instead of depending too much on other cloud companies – this saves money when working at a very large scale
Mark Zuckerberg said during the earnings call, “This is going to be a big year for delivering personal super intelligence, accelerating our business, building infrastructure for the future, and shaping how our company will work going forward.”
“We are seeing a major AI acceleration. I expect 2026 to be a year where this wave accelerates even further on several fronts. We are starting to see agents really work. This will unlock the ability to build completely new products and transform how we work,” he added.
Even with this huge spending, Meta expects to make strong profits – around $50–55 billion in free cash flow in 2026. Their advertising business (where most of their money comes from) is growing quickly, which helps pay for the AI push.
Why this matters for Meta in the AI race
Meta’s plan shows how serious the competition is between companies like Meta, OpenAI, Google, Anthropic, and xAI. Whoever builds the biggest and best AI systems first will have a big advantage.
If Meta really spends close to $135 billion in a single year, it would be one of the largest AI infrastructure investments ever made by one company.
As Meta prepares to deploy Llama 4 and integrate more advanced AI across its 3.2 billion daily active users, the 2026 spending surge positions the company to potentially leapfrog rivals in model scale, inference speed, and product innovation. Whether the massive outlay translates into sustained revenue growth and market share gains will be one of the defining questions for Big Tech in 2026.
