Meta reportedly abandoned a plan to cut staffing on some teams by as much as 60% after its AI agents failed to produce the expected gains and employees pushed back. The account, based on internal documents described by Reuters, says the effort operated under the name Project OT.
The proposal would have assigned remaining work to smaller groups of employees supervising virtual AI workers. Chief executive Mark Zuckerberg halted a second layoff wave on May 19, only hours before an initial round and months before the later cuts were due to occur, according to the report.
Employee opposition intensified amid beliefs that software recording mouse clicks and keystrokes was helping train replacements. Internal sentiment reportedly fell from 74% to 55%. Investor criticism of Meta’s large AI budget added pressure, while Zuckerberg acknowledged in July that agent technology had not advanced as quickly as expected.
The episode is a reported internal reversal, not a public product launch, and Meta’s complete planning documents are not available in the source. It nevertheless illustrates a practical constraint on aggressive automation targets: deployment schedules can move faster than the reliability of the agents expected to take over the work.