Meta Project OT: When AI-Native Redesign Outran Agent Readiness

Meta's Project OT provides an unusually detailed case of an enterprise attempting to redesign its workforce and operating model around anticipated AI-agent capabilities.

This 18-page AIGW Deep Dive examines the company's AI-native organizational experiments, including smaller AI-supported product teams, broader builder roles, flatter coordination structures and consequential restructuring proposals.

Drawing on formal Meta disclosures and Reuters' reporting on internal documents, organizational changes and employee experiences, the study distinguishes proposed changes from implemented workforce measures. Meta's May 2026 reduction affected approximately 8,000 employees, while smaller AI-supported team structures were implemented in selected units. A further broad restructuring wave was subsequently canceled.

The report examines tensions between increased software-development activity and downstream delivery, operational reliability, human supervision, management authority and employee trust. It also addresses the controversy surrounding employee activity-data collection and the limits of evidence about AI-agent performance.

Using AIGW's Workforce Architecture approach, the case identifies the risks of restructuring work and staffing around anticipated technical capabilities before human responsibilities, organizational safeguards and reliable performance measures are sufficiently established.

The report does not conclude that AI directly displaced every affected worker, that all Project OT experiments failed or that Meta abandoned its AI strategy. It shows why organizational redesign, agent readiness, workforce accounting and employee legitimacy must be evaluated together.