In a recent discussion on Lenny's Podcast, Instagram head Adam Mosseri suggested that within the next year or two, Meta might need to impose limits on how much its employees can spend on AI tokens. "I think that you can imagine, at least in a year or two … that the burn rate of a strong engineer might be the same as their salary, or their cost of employment. And in that world, you’re going to probably need to put in some caps," he explained. AI token spend refers to the cost of processing AI prompts and responses, a topic that has gained significant attention lately. Meta recently took down an internal leaderboard tracking AI token costs after projections indicated the company could face billions of dollars in expenses by 2026.

Meta is not the only company grappling with this issue. Uber, for instance, faced its own AI cost reckoning after burning through its entire 2026 AI coding budget by April. Similarly, soaring token costs led Microsoft to cancel Claude Code licenses, instead consolidating its engineers around its own Copilot CLI tool. Mosseri emphasized that AI token costs will eventually need to be managed like any other resource, drawing a parallel to payroll or operational expenditure (OpEx). "I think of it like…any other resource," he said. "I have to decide how to deploy capacity to my different teams because I have a limited number of GPUs and CPUs and storage and RAM etc. I have to decide how to deploy OpEx for labeling budgets across my teams. I have to decide how to deploy payroll for headcount across my teams."

Token budgets, Mosseri added, would follow the same logic, with caps per engineer set proportionally to the company's confidence in their ability to use the budget in a way that yields a positive return on investment. Currently, Meta does not impose any token caps on employees, but Mosseri believes such measures could become beneficial in the future. Looking ahead, he expects token costs to decrease as AI model makers engage in pricing wars to attract users to their tools over competitors. For now, Meta has managed to curb some of its token expenses by eliminating what Mosseri described as "silly things," such as the token spend leaderboard. "It’s not that hard to build a token incinerator, and that doesn’t create a lot of value," he noted.