Meta made 'mistakes' in AI workforce shift, says CEO Mark Zuckerberg

Meta CEO Mark Zuckerberg has publicly admitted the company 'made mistakes' in its sweeping, multi-billion-dollar push to reorient its workforce around Artificial Intelligence, following a May restructuring that saw 8,000 employees laid off and another 7,000 reassigned to new AI initiatives. The admission, revealed in an internal memo, signals the immense operational challenges even tech giants face in the frantic race to dominate AI. This colossal shift, underpinned by a staggering 2026 capital expenditure forecast of $125 billion to $145 billion for AI infrastructure, has prompted internal dissent over perceived forced reassignments and broadened managerial oversight, with some teams reporting a 50:1 employee-to-manager ratio. Meta's aggressive investment reflects a broader, disruptive trend across Big Tech, as companies like Cisco, Dell, and Coinbase also cite AI as a primary driver for major workforce reductions and strategic pivots in 2026, impacting tens of thousands of jobs across the sector. However, this heavy spending is raising investor concerns about the return on investment (ROI), even as Meta's core advertising business sees AI-driven improvements. While Zuckerberg has pledged no further company-wide layoffs in 2026, Meta is now focused on finding suitable roles for reassigned staff and increasing investment in team-building, including a July hackathon, to foster cross-team collaboration on its latest AI models. Concurrently, Meta is channeling $115 million into 'America's Workforce Academies' to train workers for the construction of its burgeoning AI data centers, underscoring the long-term, structural changes AI is imposing on both corporate strategy and the wider labor market. The coming quarters will reveal whether these organizational adjustments and massive capital outlays begin to translate into tangible, monetizable AI products that reassure a skeptical market.