Word on the street is that on May 18, local U.S. time, Meta dropped an internal memo on employees, finally revealing the big layoff and company shake-up happening this week.
According to the memo, Meta is set to fire 10% of its people globally on Wednesday, May 20, and they’re already hinting at even bigger layoffs later in the year.
Meta’s own financial reports say that by the end of March, Meta had around 78,000 workers, so 10% comes out to about 7,800 folks.
The memo says the pink slips will go out in three waves around the world,and folks in North America got the heads-up to work from home on Wednesday.
Way back on April 23, local time, in a memo to employees,Metaalready said it was looking to cut10%of the workforce to become more efficient and cover the huge bills from AI investments. That same memo also said the cuts would happen onMay20.
But it’s not just about firing people – Meta is also shaking up its whole organizational structure, cutting down tons of middle management to make things flatter. Meta’s Chief People Officer, Janelle Gale, pointed out in the memo that a bunch of department leaders are now using “AI-native design principles” when rebuilding their teams. The idea is to move to smaller, more empowered teams that can move faster.
In this so-called “all-in on AI” shake-up, it’s not only about people losing their jobs – Meta is also planning to shift as many as 7,000 employees over to new AI-focused projects. Some of those moves are already happening.
If you add up the layoffs and the role changes, this whole reorganization is hitting roughly 20% of Meta’s employees. Plus, the company has pulled the plug on another 6,000 job openings that were on the market.
This whole thing is just part of Meta’s big reorganization plan for this year. They’re pouring money into AI, trying to weave AI agents into everything – from their products to how they work internally.
Back in December, word from insiders (via Cailianshe) said that Meta was cooking up a cost-cutting plan for fiscal 2026, looking to slash the metaverse budget by as much as 30%. That would mean even more layoffs, but the exact details are still being hammered out.
Just recently, on April 22, 2026, an internal memo leaked that Meta wants to put new tracking software on its U.S. employees’ work computers. It’ll track mouse moves, clicks, and even keystrokes – all to train their AI models. The tool will work across various work apps and websites.
And back in March,Metafounder and CEO Mark·Zuckerbergsaid he wants everyone – inside and outside the company – to eventually have their own personal AI buddy. He’s starting with himself. Rumor has it he’s building a CEO AI agent to help him do his job and grab info faster.
In that same month, Meta rolled out a new trick on Facebook Marketplace, powered by Meta AI. It lets users throw up a listing in one click, and the AI helps answer buyer questions. Just upload a pic, and Meta AI auto-creates a draft listing, fills in the deets, and even suggests a price based on what similar stuff sells for in your area. Plus, it shows your past listings, what you’ve sold, and your seller rating right at the top of your Marketplace profile.
On April 29, local time, Meta dropped its Q1 earnings: they pulled in 56.31 billion in revenue, up 33% from last year.Of that, ad money brought in55.02 billion, also up 33%; operating profit hit 22.87 billion, up 30%; earnings per share came in at 10.44, up from 6.43 a year ago. And they’re forecasting Q2 revenue between 58 and 61 billion.
And here’s the kicker: Metaupped its full-year capex forecast – from115billion to135billion, now bumped up to125billion to145billion. That’s about10billion more than just three months ago, and it’s the highest ever. In just one quarter, they spent19.84billion on capex, way up from last year, showing they’re all in onAIand data centers.
Meanwhile, Meta’s daily active users actually shrank for the first time ever from one quarter to the next – down from 3.58 billion to 3.56 billion.
This past April, Meta inked a multi-billion-dollar deal with Amazon to rent their Graviton chips for AI work. And back in February, Meta teamed up with Nvidia in a big strategic partnership that spans on-prem, cloud, and AI infra. Oh, and Meta is the first company to go all-in on Nvidia’s Grace CPU on a big scale.
Here’s the breakdown: Meta is building massive data centers for AI training and inference to back up its long-term AI plans. With this partnership, they’ll be rolling out Nvidia CPUs, millions of Blackwell and Rubin chip GPUs, and weaving Nvidia’s Spectrum-X™ Ethernet switches into Meta’s Facebook open switching system.
