The Meta Platforms Inc. pavilion ahead of the World Economic Forum (WEF) in Davos, Switzerland, on Saturday, Jan. 19, 2025. The annual Davos gathering of political leaders, top executives and celebrities runs from January 20 to 24. Photographer: Hollie Adams/Bloomberg via Getty Images
- Meta's 6,500-member Applied AI unit is facing severe internal backlash, with employees reporting low morale and forced reassignments.
- Staff are protesting both the repetitive nature of their AI-training tasks and a controversial keystroke-monitoring program.
- CEO Mark Zuckerberg has acknowledged the internal distress and admitted to organizational errors, promising a path toward resolution.
Workplace Unrest Brews Within Meta’s AI Division
Meta, the tech giant driving hard into the artificial intelligence race, is currently facing significant internal turbulence. Reports indicate that the company’s newly formed Applied AI unit—a massive team of approximately 6,500 engineers and product managers—is grappling with severe employee dissatisfaction, with some staff members going as far as to describe their working conditions as a “soul-crushing gulag.”
The “Draftee” Phenomenon
Tensions reached a breaking point during a recent internal livestream. The presentation was interrupted by an explicit outburst from an attendee who demanded that a senior executive be confronted regarding the deteriorating work culture. This incident is widely seen as a manifestation of the mounting frustration among personnel who claim they were involuntarily reassigned to the Applied AI division.
Employees have labeled themselves as “draftees,” alleging they were given an ultimatum: join the new AI project or face departure from the company. The nature of the work, which largely involves the repetitive generation of coding puzzles and training datasets, has been a significant point of contention for highly skilled engineers who feel their talents are being misaligned with their professional expectations.
Key Points of Contention
- Mandatory Reassignments: Thousands of staff report being “drafted” into the Applied AI unit without professional agency.
- Monotonous Workflows: Engineers cite the task of creating AI training data and puzzles as monotonous and unrewarding.
- Employee Surveillance: A petition signed by over 1,600 Meta employees is protesting a new internal program designed to track keystrokes and clicks to generate training data for AI models.
Leadership Response and the Road Ahead
The internal atmosphere has become so strained that even Meta’s Chief Product Officer, Chris Cox, described the environment as “brutal” during a recent company-wide call. Recognizing the escalating unrest, CEO Mark Zuckerberg issued an internal memo admitting that recent organizational shifts had caused significant distress. Zuckerberg acknowledged that management had made mistakes in the transition and committed to rectifying the situation, reiterating that Meta’s goal remains to be the premier destination for top-tier global talent.
As the company continues to funnel billions into its AI ambitions, the challenge remains: can Meta reconcile its aggressive “AI-first” roadmap with the morale and autonomy of its workforce? With thousands of employees protesting surveillance and reporting burnout, the stability of Meta’s long-term AI strategy hinges on its ability to mend this fractured internal culture.