Meta’s AI overhaul runs into worker backlash and rising costs

Meta’s attempt to reorganize its workforce around artificial intelligence has run into employee resistance, rising compensation costs and questions over whether its AI investment is delivering the expected gains, according to Reuters.

The company’s employee count is shrinking while revenue per worker rises. But Meta spent just over $30 billion on total worker compensation in the first half of 2026, including healthcare and benefits. Excluding severance, compensation was nearly 30% higher than in the first half of 2025.

Meta’s data-center spending is also accelerating. Capital expenditure is expected to reach nearly $170 billion next year, according to LSEG data, potentially increasing depreciation and amortization costs as the company seeks savings from a smaller workforce.

Reuters reported that Meta had considered cutting as much as 60% of its workforce under a restructuring plan before scaling the proposal back. The company ultimately laid off about 10% of employees, while plans for a second wave of restructuring were canceled.

The restructuring has also created operational concerns. Changes to Meta’s codebase have expanded faster than user-facing features, while the reliability and security of automated work remain uncertain, Reuters reported. Plans to organize employees into small teams focused on top-performing talent also fueled confusion and resentment.

Meta’s experience reflects the broader difficulty of redesigning large organizations around new technology. The company is now testing whether higher AI spending and a smaller workforce can translate into lasting productivity gains.