In spring 2026, Meta reassigned between 6,500 and 7,000 engineers to its applied AI unit, triggering a morale collapse with employee favorable ratings dropping from 74% to 55%. The company then laid off approximately 10% of its workforce, cutting 8,000 positions after considering a 60% reduction under Project OT. The Model Capability Initiative, which tracked employee keystrokes, gathered over 1,600 signatures against what staff called an intrusive surveillance program. Meta has invested between $125 billion and $145 billion annually in AI infrastructure, but internal code improvements reached 220% versus only 36% for user-facing features. Yann LeCun, Meta’s head of AI research, departed the company at the end of 2025.
Source: Read the original article
