According to news released by The Information, social media group Meta is taking measures to limit the use of artificial intelligence by internal employees, because the current usage is too high and is expected to generate billions of dollars in usage fees. This is also a very high cost for Meta, so Meta must take measures to control to avoid the cost of AI usage getting out of control.
Employees have been encouraged to use AI and use usage as an assessment indicator:
Previously, Meta was actively promoting employees to use artificial intelligence-related tools to improve efficiency. To encourage employees to use artificial intelligence, Meta even included AI usage in some performance appraisals. Of course, as long as there is an appraisal, there will definitely be cheating methods. For example, there are posts on Reddit discussing how to quickly consume a large number of Tokens.
Now Meta is beginning to shift to a Tokens minimization usage strategy, that is, AI should not be used for unnecessary tasks to reduce usage. To this end, Meta will set an AI budget cap for each team, establish a Token usage cap, and encourage employees to use their own artificial intelligence tools such as Meta Code, etc., to control the surge in AI costs.
It is reported that Meta employees have consumed up to 60 trillion Tokens in the last 30 days alone. AI costs have become a huge expense for Meta. Meta’s expenditure on AI usage is expected to reach billions of dollars throughout 2026. Even a company of Meta’s size believes that the cost of AI use is too high and must be controlled.
Develop tools to track employee AI usage:
Meta plans to start real-time employee AI usage tracking and set preliminary limits in 2026, such as the maximum AI quota each team can obtain. By 2027, Meta will have dedicated structured management, including strict budget allocation and tool support. This also includes the development of an internal platform to track employee AI usage and related cost expenditures in real time.
Of course, it is not just Meta that is controlling the cost of AI use. Other technology companies such as Uber, Microsoft, and Salesforce have recently implemented similar restrictions. Companies have begun to shift from encouraging employees and unlimited use to rational management and control to control costs. If unrestricted use continues, the cost of AI use may become a major burden on business operations. It is difficult to say whether efficiency can be improved, but huge expenditures are certain, so AI usage can only be controlled.