Meta is ending a year-long experiment that tied employee performance reviews to how heavily they used AI tools. In an internal announcement this week, the company told workers their evaluations would no longer depend on AI usage, according to three employees who received the message.
Almost a year ago Meta said staff would be graded on “AI-driven impact,” which in practice meant how much they leaned on chatbots and agents, with workers labeled “AI Native,” “AI First,” or “AI Enabled” based on usage, Wired reported.
The change comes as employees begin testing Hatch, an internal agentic AI tool that Meta describes as its most advanced AI project yet. Workers who received the memo said token consumption is still surging as Hatch pilots expand, even with the usage quota gone.
The reversal removes an incentive that rewarded burning tokens over actual productivity, and it leaves Meta betting that organic adoption of Hatch will embed AI in daily work more effectively than any grading scheme did.
