Overview
- Meta told employees this week that performance reviews will no longer use raw AI usage or token counts to judge impact.
- Workers report the company is still encouraging testing of Hatch, an internal agentic AI on corporate devices that consumes far more tokens than standard chatbots.
- A previously disclosed project that captured keystrokes and device activity to train models was paused after employee backlash, and that episode has left many staff wary of linking personal accounts to agents.
- Employees say cultural and managerial pressure to adopt AI tools persists despite the review change, so token consumption and informal incentives to use agents remain high.
- Some staff fear successful deployment of agents could trigger more job cuts, and Reuters reporting shows planned additional layoffs were scaled back after AI‑related software bugs and slower agent progress.