Meta explored massive workforce reductions across multiple teams to become "AI native," but reversed course after employee pushback and data showed AI agents were ineffective, internal documents reveal.
Meta considered eliminating as much as 60% of staff in certain divisions through two planned reduction waves, according to internal documents reviewed by Reuters. The cuts were part of a broader strategy to restructure the company around artificial intelligence capabilities.
The plan did not proceed. Two factors derailed the initiative: significant staff resistance and empirical evidence that AI agents failed to deliver promised results.
The company's pivot reflects broader challenges in the AI sector. While large language models and generative AI have captured investor attention and corporate investment, autonomous AI agents remain largely unproven at scale. Meta's internal data suggested the technology was not ready to replace human workers at the proposed scale.
The reversal also highlights tensions between corporate efficiency goals and workforce concerns. Meta has already conducted layoffs in recent years, reducing headcount significantly. A 60% cut in multiple teams would have represented a dramatic escalation that triggered substantial internal opposition.
Mark Zuckerberg has emphasized AI as central to Meta's future strategy, including the development of AI agents that could eventually assist or replace certain functions. The company continues investing heavily in AI infrastructure and research. However, this investigation indicates executives faced practical limits on how aggressively they could restructure around unproven technology.
The episode underscores a gap between AI hype and operational reality. While companies race to integrate AI, the technology's actual capability to perform complex human work remains constrained by current model limitations, hallucination issues, and reliability concerns.
Meta did not immediately respond to requests for comment on the investigation.
Malaysia's Communications Minister Fahmi Fadzil announced plans to regulate e-commerce platforms through a registration mechanism, responding to complaints from users and merchants.
Europe, India, Brazil, and China are attempting to challenge Big Tech dominance, but their efforts continue to face significant obstacles, according to analysis from "The Rulers: Corporate Power in the Age of AI and the Cloud."
Adam Mosseri testified Tuesday that Instagram did not intentionally conceal information about low usage of teen safety features. The testimony marks the first time a top Meta executive has taken the stand in a lawsuit filed by 29 US states.
X has forced Nitter, an open-source alternative frontend, offline by sending cease-and-desist letters. The project allowed users to read X posts without logging in or using the app.