AI Token Costs on the Rise
Meta is bracing itself for the high costs associated with AI experimentation, with Instagram head Adam Mosseri predicting that token budgets will need to be managed like any other resource.
In a recent interview, Mosseri stated that he can see a time in the future, perhaps only a year or two, when putting limits on Meta employees' AI token spend will become necessary. This move comes as the company grapples with the soaring costs of processing AI prompts and responses. Meta has already shut down an internal AI token spend leaderboard after AI costs put the company on track for billions of dollars in 2026.
Comparing AI Token Costs to Other Resources
Mosseri believes that AI token costs will have to be managed just like any other resource, offering an analogy to things like payroll or operating expenditure (OpEx), which is the day-to-day costs of running a business. He explained that token budgets will be the same as these other resources, noting that the cap per engineer would have to be proportional to the company's trust in their ability to use the budget in an ROI-positive way.
A Shift in AI Model Pricing
Further down the road, Mosseri expects token costs to come down as the AI model makers enter a pricing war to attract people to use their tools over their competitors. However, for now, the company has managed to rein in its token costs by shutting down the "silly things" that it was doing, like the token spend leaderboard.
"It's not that hard to build a token incinerator, and that doesn't create a lot of value," Mosseri said.
A New Era of AI Cost Management
As companies like Uber and Microsoft also grapple with the high costs of AI experimentation, Mosseri's comments highlight the need for a new approach to AI cost management. By comparing AI token costs to other resources, Meta is setting the stage for a more sustainable and responsible approach to AI adoption.
- Uber blew through its 2026 AI coding budget by April.
- Microsoft canceled Claude Code licenses, consolidating its engineers around its own Copilot CLI tool instead.