Zing Open Zing
Technology

Meta Fires 8,000 Workers After Using Their Data For AI Training

26 May 2026 · Source balance: Left of Centre

On May 20, 2026, Meta laid off approximately 8,000 employees—roughly 10 percent of its workforce—while also canceling 6,000 open positions. The layoffs followed an April 23 announcement that gave employees a month of uncertainty about who would be affected. During this period, leaked audio from an April 30 all-hands meeting showed CEO Mark Zuckerberg explaining that Meta's AI models were being trained by observing employee work patterns through keystroke, mouse movement, and screenshot tracking. The company framed the layoffs as necessary for competing in artificial intelligence, with Zuckerberg stating that AI is 'the most consequential technology of our lifetimes' and that 'success isn't a given.' Meta is investing between 125 and 145 billion dollars in AI infrastructure in 2026. The moves have sparked employee resistance, with over 1,500 workers signing an internal petition objecting to the surveillance program, citing concerns about privacy, consent, and workplace trust. The company defends the tracking as essential for training AI agents, while also reassigning 7,000 employees to AI-focused roles.

Key facts

  • Meta laid off 8,000 employees on May 20, 2026, representing 10 percent of its 78,000-person workforce
  • The company announced layoffs on April 23 but did not identify affected employees until May 20, creating what staff called 28 days of hell
  • Meta's Model Capability Initiative tracks keystrokes, mouse movements, and screenshots to train AI models; over 1,500 employees petitioned against it
  • Meta is investing between 125 and 145 billion dollars in AI infrastructure in 2026, nearly double its 2025 spending
  • Leaked audio from April 30 all-hands meeting shows Zuckerberg explaining that Meta trains AI by observing smart people work; the audio remains unverified
  • Meta also canceled 6,000 open positions and reassigned 7,000 employees to AI-focused divisions
  • The company states employee data is not used for surveillance or performance tracking, only AI training; Zuckerberg acknowledged communication could have been better

How different outlets framed it

Left-leaning coverage: Left-leaning outlets like Kotaku and Futurism frame the situation as inherently exploitative and cruel. They characterize Zuckerberg as deliberately 'taunting' employees by boasting about using their data to train AI that will replace them, while they await layoff notifications. They emphasize the 'sadistic' announcement of a month-long waiting period and highlight the contradiction of praising employee 'intelligence' while simultaneously firing them. This framing focuses on worker vulnerability, data extraction without consent, and what they see as a callous disregard for employee welfare.

Right-leaning coverage: Right-leaning or business-focused sources frame this as a strategic necessity for competitiveness. They highlight that Meta faces 'fierce competition' in AI and that companies like Google, Amazon, and Microsoft are similarly restructuring. These sources emphasize Zuckerberg's stated goal of repositioning Meta as an AI leader and note that the company is making calculated bets on automation. They discuss whether the strategy will pay off and present layoffs as a difficult but rational reallocation of human capital from 'legacy products where innovation has plateaued' to AI where 'competitive differentiation remains open.'

Limits of this briefing

The leaked audio recording featuring Zuckerberg remains unverified and unauthenticated; Meta did not confirm its authenticity. Multiple sources note the recording could not be independently verified as genuine or ruled out as synthetically generated, though its contents align with Meta's public statements. Coverage heavily emphasizes the employee and critic perspective; fewer right-leaning sources provided substantive business defenses.

Sources

More stories

See all latest stories