Meta is considering imposing a cap on its engineers' use of artificial intelligence due to cost.

 

Artificial intelligence
Meta 

Meta is considering imposing a cap on its engineers' use of artificial intelligence due to cost.

Adam Mosseri, head of Instagram, revealed that Meta may have to impose limits on its employees' consumption of AI tokens within the next two years, due to the rapidly rising costs of running intelligent models.

During his appearance on Lenny's Podcast, Mosseri said he expects a top engineer's spending on AI tools to reach a level equivalent to their salary or the cost of hiring them, which would prompt the company to set a cap on usage.

He added: “You can imagine that in just one or two years, an engineer’s spending on artificial intelligence could equal the cost of hiring him. At that point, it would make sense to impose limits on its use,” according to a report published by TechCrunch and reviewed by Al Arabiya Business.

The costs of artificial intelligence are causing concern for companies.

The term AI Token Spend refers to the cost of processing requests and responses within AI models, and is one item that has seen a significant increase as companies increasingly rely on AI-based programming and productivity tools.
Meta recently suspended an internal dashboard that displayed employees' rankings based on their token consumption, after data showed that spending was on track to reach billions of dollars by 2026 if the current pace continued.

It's not just "Meta"

Meta is not the only company reconsidering its AI spending; Uber faced a similar crisis after exhausting its 2026 budget for AI-based programming tools by April.
Microsoft was also forced to cancel Cloud Code licenses for some of its engineering teams due to high usage costs, directing developers to rely more on its own Copilot CLI tool.

Managing codes such as payroll and resources

Mosseri believes that AI code will become a resource that must be managed in the same way as other resources within companies, such as salaries, operating budgets, or infrastructure.
He explained that management already has to allocate available resources, whether it be graphics processing units (GPUs), processors, memory, or operating budgets, and therefore budgets for using artificial intelligence will be subject to the same logic.
He noted that any future cap on token use would be linked to the company’s confidence in the employee’s ability to generate a positive return on investment (ROI) through their use of these resources.

There are no restrictions at the moment, but prices may decrease.

The head of Instagram confirmed that Meta currently does not impose any limits on the use of AI codes within the company, but believes that implementing such a policy could be beneficial in the future as costs continue to rise.

Conversely, he predicted that the prices for using artificial intelligence models would decrease in the long term, as competition intensifies among companies developing the models, and they seek to attract more customers by lowering prices.

Stop unhelpful ideas to reduce spending

Mosseri noted that Meta had already begun to control part of its expenses by halting some internal projects that he described as "unhelpful ideas," including a dashboard ranking employees by their consumption of AI tokens.
He concluded by saying, "It's not difficult to build a tool that burns huge amounts of code, but that doesn't create real value."


Post a Comment

0 Comments