Inside the 45,000-Job Axing That Wasn’t: How Meta’s Push for an ‘AI-Native’ Workforce Has Been Dumped Now

Fresh off agreeing to pay a historic $17 billion settlement in a California federal lawsuit over youth social media addiction and child safety, Meta Platforms is confronting an internal crisis over its operational core: the near-total collapse of a secret, corporate-wide effort to automate its human workforce using artificial intelligence.

According to internal company documents and reporting on executive operations, Chief Executive Officer Mark Zuckerberg had quietly set in motion a restructuring scheme code-named Project OT (Organisation Transformation). Hatched during an executive retreat at Zuckerberg’s Hawaii compound in early 2026, the blueprint envisioned a radically restructured "AI-native" enterprise, an investigation reported by Reuters claims.

Had it been fully executed, Project OT would have led to the layoff of roughly 45,000 workers—marking what would have been the single largest AI-driven job displacement in human history.

The Hawaii Blueprint: The Rise of 'Builder Pods'

Inspired by high-velocity artificial intelligence startups operating across East Asia, senior Meta executives—including Chief Data Officer Alex Schultz and Head of Product Naomi Gleit—sought to dismantle traditional tech-company hierarchy.

Under the proposed model, standard corporate teams of 10 to 20 engineers, product managers, and specialists were to be replaced by streamlined "builder pods" consisting of just three to five human workers. These small pods were designed to oversee fleets of autonomous AI software agents tasked with writing code, analysing data metrics, managing product pipelines, and handling day-to-day administrative duties.

To achieve this flattened organizational matrix:

• Severe Team-Level Reductions: Scenario planning mapped out workforce reductions of up to 60% across targeted product, business, and administrative divisions.

• Massive Scope: Applied across Meta’s global workforce of nearly 78,000 employees, the targeted cuts projected an unprecedented elimination of upwards of 45,000 positions over two major layoff waves planned for May and November.

• Management Eradication: Middle management was slated for near-total removal, with single unit directors designated to oversee 30 to 50 independent builder pods simultaneously.

The Implosion: Hallucinations, Bottlenecks, and Mutiny

The plan began to unravel when deployed autonomous AI models proved incapable of operating at the level required for enterprise-scale engineering and operations.

While the algorithms handled basic scripting efficiently, internal performance audits revealed severe operational bottlenecks. The AI models struggled with open-ended problem solving, frequently outputted faulty code, and exhibited "spiky" intelligence that routinely required human intervention to correct errors. Rather than boosting productivity, human workers spent double the time auditing and debugging AI output.

Simultaneously, word of Project OT leaked internally. Morale plunged across the organisation as employees realised the AI tools they were being instructed to integrate were engineered specifically to render their positions obsolete.

As internal friction mounted and AI capacity targets fell drastically short, investor scrutiny over ballooning capital expenditures forced leadership to evaluate the operational risks of gutting human oversight. Hours before the first major reduction wave was set to take effect, Zuckerberg abandoned the second, deeper wave of cuts scheduled for late 2026.

A Staggering Scale in Context

While Meta eventually proceeded with a smaller, 10% restructuring wave in May 2026—affecting roughly 8,000 staff alongside the redeployment of 7,000 others into core AI research—the aborted 45,000-person cut stands as a stark milestone.

Had Meta carried out the full 45,000-person axing, it would have shattered every historic benchmark in the tech sector, dwarfing Meta’s own 2022 cuts of 11,000 workers and Amazon's multi-wave reductions of 27,000.

Across broader corporate history, a 45,000-person termination wave would have ranked behind only IBM’s 60,000-person pivot in 1993 and Sears' 50,000-person catalog shutdown, instantly cementing itself as the largest automation-driven displacement of human labor on record.

Meta acknowledged that Project OT served as an internal organizational study aimed at shifting resources toward artificial intelligence infrastructure, but maintained that the 60% reduction figures were theoretical planning bounds rather than finalized enterprise-wide targets.

Coming alongside a historic legal battle in California that fundamentally reshapes how Meta designs its consumer products, the quiet collapse of Project OT underlines a sobering reality for Silicon Valley: replacing human capital with autonomous AI remains far easier on a whiteboard than inside a global corporation.