Meta is preparing to cut roughly 8,000 jobs, or about 10% of its workforce, as the company tries to keep a lid on costs while pouring more money into AI. The move is blunt, familiar, and very much in character for Big Tech right now: spend aggressively on models and infrastructure, then trim headcount to keep investors calm.
The timing says a lot. Meta has already warned that capital expenditures will jump by at least 60% this year compared with 2024, driven by investment in Meta Superintelligence Labs and its core business. At the same time, free cash flow is expected to fall 83% year over year, which is the kind of number that makes even growth-hungry shareholders start checking the exits.
Meta’s AI bill is getting expensive fast
This is not a company improvising for the first time. Meta has already gone through more than 20,000 layoffs in earlier rounds, and those cuts helped kick off its efficiency push in 2022 and 2023. The new round looks less like a retreat than a continuation of the same playbook: keep the AI spending binge, but make the workforce footnote smaller.
That strategy is spreading across the sector. Amazon says it will cut around 16,000 workers this year, Block is reducing its staff by around 4,000, Salesforce has announced roughly 1,000 cuts tied to AI automation, Snap is trimming around 1,000 jobs, and Microsoft said it would offer buyouts to 7% of staff. Translation: Wall Street likes AI, but it likes margins too.
Big Tech’s new balance sheet math
The pressure is obvious. Training and running AI systems costs real money, and the biggest platforms are now competing to show they can fund the race without turning profits into a horror story. The layoffs are doing double duty: cutting payroll now while sending a signal that management is willing to be ruthless about efficiency.
What makes Meta especially interesting is that it is both a spender and a precedent. Once a company with its scale starts treating layoffs as a routine financial instrument, smaller tech firms tend to follow the script with even less ceremony. Expect more of that if AI costs keep rising faster than revenue can justify.
A new data push raises fresh questions
The layoffs also arrive alongside a separate report that Meta plans to record employees’ keystrokes to collect training data for its AI models. If true, that would underline how far companies are willing to go to make systems behave more like people at keyboards, and how little patience they have for waste anywhere else in the organization.
The bigger question is whether this is the start of a cleaner, leaner AI era or just a temporary truce with costs. Meta seems to believe it can buy enough time with cuts and enough growth with AI to please investors. That bet could work for a while. It could also get very expensive, very quickly, if the revenue story does not catch up.