Companies pull back on AI as Meta scraps staff-cut plan

The Guardian reports that employers around the globe are walking back AI rollouts, after staff ended up spending more hours cleaning up errors and low-quality output than the tools saved. The clearest case is Meta, which has scaled back plans to axe staff and lean on AI agents.

Meta's shelved restructuring, codenamed Project OT for 'organization transformation,' had planned to shrink some teams by as much as 60% and hand their day-to-day work to autonomous AI agents overseen by smaller 'builder' squads. An internal document framing the plan described a workplace where "AI-ready tools and agents interact, workflows are automated, [and] new builds are AI-first." The night before the first layoff wave landed, on May 19, Mark Zuckerberg scrapped a planned November follow-up. Code changes to internal platforms rose 220% year over year, but changes that reached users rose only 36%; major technical and security incidents climbed 40%, and the time staff spent firefighting them was up 70%. Internal posts described AI agents taking "large-scale, disruptive actions that humans are unlikely to execute."

The wider data the piece cites is no kinder. A survey it quotes found 70% of managers had seen direct reports commit AI-related errors with real business consequences, some causing financial losses above $50,000. Forrester Research, in its 2026 'Future of Work' report, put the share of employers who regret AI-related layoffs at 55%.