1 hour ago
27
Meta is spending so much on AI that it had to tell employees to stop using other people’s AI. The company is aggressively pushing its workforce toward proprietary tools like MetaCode while pulling back from third-party coding assistants, a move driven by a monthly AI bill that would make most CFOs physically ill.
Here’s the number that explains the urgency: Meta employees consumed 73.7 trillion AI tokens in a single recent month. At enterprise pricing, that translates to roughly $221 million per month, or about $2.65 billion annually.
The internal pivot and what it actually looks like
Rather than simply telling employees to use less AI, Meta is channeling usage toward its own tools. The company has developed MetaCode as its primary internal coding assistant and is making it clear that employees should use it instead of external alternatives.
This isn’t a suggestion. AI tool adoption metrics are now factored into employee performance evaluations. In English: how much you use Meta’s AI tools could affect your review, your rating, and presumably your continued employment at the company.
Engineering teams have been given specific targets for AI-assisted code generation. The goal is for 65% to 75% of committed code to be generated through AI tools by the first half of 2026. Some teams are aiming even higher, at 80%.
To keep costs from spiraling further, Meta has implemented spending caps and real-time usage trackers. The company has also reduced its reliance on external vendors, which is both a cost-cutting measure and a way to keep proprietary data inside Meta’s walls.
The bigger infrastructure bet
The push toward in-house AI tools is just one piece of a much larger investment strategy. Meta’s total AI infrastructure spending estimates are staggering, with projections suggesting expenditures in the range of $125 billion to $145 billion. By 2028, Meta’s data center commitments alone could total $600 billion.
CEO Mark Zuckerberg has positioned AI as the cornerstone of Meta’s future corporate strategy. The company has been integrating generative AI technologies across its platforms, including Facebook, Instagram, and WhatsApp. The internal tool mandate ensures that the data generated by employee usage feeds back into Meta’s own model training rather than enriching competitors.
Layoffs, AI, and the resource reallocation
This AI-first strategy comes alongside significant workforce reductions. Meta has announced layoffs of around 8,000 employees, approximately 10% of its workforce. The message is fairly transparent: fewer humans, more AI, and the humans who remain should be using AI as aggressively as possible.
Tying AI adoption to performance reviews is a corporate culture shift that few other companies have been willing to formalize so directly.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.
English (US) ·