Meta announced in its new performance evaluation guidelines that it would not evaluate employees based on how often they use AI tools. Instead of AI usage rates, the company will focus on the tangible impact of the work employees produce.
Last year, the “AI-focused impact” criterion led to competition among employees centered on increasing usage statistics, with titles such as “Token Legend” being used. Meta informed engineering teams that AI usage dashboards and token counts would no longer be performance metrics. The company is also continuing to test an AI agent called “Hatch” that can perform operations on computers.
Why it matters
The change puts the work outcomes produced with AI tools at the forefront instead of trying to show that employees are using those tools more. As a result, the direction of the incentives in performance evaluations is changing, particularly for engineering teams; competition based on usage dashboards and token counts no longer has value as a metric. The decision’s arrival after titles and behaviors based on increasing AI usage statistics emerged last year shows that the company acknowledges that quantitative indicators do not explain employee performance on their own. Meta’s continued testing of its Hatch agent, which can perform operations on computers, shows that AI’s role within the company has not diminished, and that the discussion has shifted from the amount of usage to the impact produced by that usage.
Background
Meta is not a new name in the FikirPilot archive: we have published 5 news reports mentioning this name in the last 90 days; the most recent is dated September 2, 2026.