Meta drew up plans to cut some teams by 60 percent in a push to run on AI
An internal project called 'Organisation Transformation' mapped out deep headcount reductions. Meta confirmed the plan existed but has not said which teams were in scope.
Key points
- Meta created an internal plan, codenamed Project OT, to cut some team headcounts by as much as 60 percent, according to Reuters.
- The plan called for two separate rounds of layoffs as part of a wider push to make Meta "AI native".
- Meta confirmed Project OT existed but declined to say which teams it covered.
- The plan appears to have been shelved; Reuters described it as something Meta "created" earlier this year, not something currently active.
Meta spent part of this year developing a detailed plan to reduce some of its workforce teams by up to 60 percent, Reuters reported, citing two people familiar with the company's internal workings.
The plan had a name: Project OT, short for Organisation Transformation. Its goal was to make Meta "AI native", meaning a company whose core work is carried out by AI systems rather than the humans who do those jobs today.
What exactly did Meta plan to do?
Project OT mapped out two separate rounds of layoffs. The severity varied by team, but the most extreme scenarios reached 60 percent of headcount. That's a gut-punch figure for any department.
Meta confirmed the plan to Reuters. It wouldn't say which teams were targeted, or whether any cuts under this project have already happened. That silence is doing a lot of work.
Internal strategy documents float around large tech companies constantly and never become policy. Meta's willingness to acknowledge this one suggests it was serious planning, not a whiteboard sketch.
What does "AI native" actually mean for workers?
It means replacing human labour with AI systems, not just layering AI tools on top of existing jobs. The distance between those two things is enormous.
Adding an AI tool might help a marketing team write faster. Going "AI native" in that same team could mean deciding the work needs three people instead of thirty, because software handles the rest. Workers in roles built around repeatable, structured tasks should pay attention to which parts of their work AI already handles well at other companies.
We first covered Project OT on 26 August 2026, around the same time we reported on how IBM and OpenAI are retraining tens of thousands of consultants for exactly this kind of AI-forward shift. The pressure on white-collar headcount is coming from more than one direction.
Should employees at big tech companies be worried?
This is worth watching closely. Meta's plan, as described, was exploratory, and the company hasn't confirmed it triggered specific job losses.
But the direction is clear. When a company Meta's size formally models 60 percent reductions in some teams, that modelling shapes real decisions eventually, even if the first version gets shelved. The skills worth building now are the ones AI handles poorly: judgment, accountability, context, relationships.
Common questions
Did Meta actually carry out these layoffs?
Meta hasn't confirmed any layoffs directly tied to Project OT. The plan appears to have been a scenario-planning exercise; whether it led to cuts in specific teams isn't yet known.
What does it mean for a company to go "AI native"?
It means building the company around AI doing the core work, rather than humans using AI as a supporting tool. In practice, it usually means fewer employees doing tasks that software can now handle.
