During a recent interview, Adam Mosseri, head of Instagram, indicated that a future scenario, potentially within one to two years, would necessitate implementing limits on Meta employees' expenditure on AI tokens.
"I think that you can imagine, at least in a year or two," Mosseri elaborated on Lenny's Podcast, "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."
The concept of AI token spend, which refers to the computational costs associated with processing artificial intelligence prompts and generating responses, has garnered significant attention recently. Meta, for instance, discontinued an internal leaderboard tracking AI token expenditure after projections indicated the company was headed towards billions of dollars in AI-related costs by 2026.
Meta's reconsideration of its AI experimentation strategy is not an isolated incident. Uber similarly faced an "AI reckoning" after exhausting its entire 2026 AI coding budget by April. Escalating token costs also prompted Microsoft to cancel its Claude Code licenses, opting instead to centralize its engineers' efforts around its proprietary Copilot CLI tool.
Mosseri clarified his perspective, stating that AI token costs must be managed akin to any other corporate resource. He drew an analogy to established financial categories such as payroll or operating expenditure (OpEx), which encompass the routine costs of business operations.
"I think of it like...any other resource," Mosseri affirmed. He continued, "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."
He further elaborated that token budgets would operate similarly, with individual engineer caps needing to be commensurate with the company's confidence in their capacity to utilize the budget in an "ROI-positive" manner.
Currently, Meta does not impose token caps on any employees, Mosseri confirmed, yet he anticipates that such limits could prove beneficial in the long term. Looking ahead, he foresees a reduction in token costs as AI model developers engage in a competitive pricing struggle to draw users away from rival platforms.
Presently, Mosseri observed, the company has successfully curtailed some of its token expenditures by discontinuing "silly things" it was undertaking, citing the aforementioned token spend leaderboard as an example.
"It's not that hard to build a token incinerator, and that doesn't create a lot of value," he concluded.
The Editorial Staff at AIChief is a team of professional content writers with extensive experience in AI and marketing. Founded in 2025, AIChief has quickly grown into the largest free AI resource hub in the industry.