Copy article

Meta US staff put up secret posters in bathroom cubicals protesting about intrusive AI "bossware"

ended 14. May 2026

Reuters reports it has seen materials asking "Don't want to work at the Employee Data Extraction Factory?" as a follow up to a report published on 21 April that Meta were to start capturing employee mouse movements, keystrokes for AI training data. 

The action comes roughly a week before Meta is expected to begin notifying the approximately 8,000 employees caught in its first round of layoffs. It is the most visible sign yet of a US labour movement taking shape inside one of Silicon Valley's giants. 

Reuters reported Meta's Model Capability Initiative (MCI) captures every keystroke, mouse movement, click, and periodic screenshot across hundreds of applications on US employees' work computers. CTO Andrew Bosworth confirmed there is no way to opt out. European employees are exempt because GDPR won't allow it. 

The employees posting flyers cite the National Labor Relations Act and direct colleagues to a petition demanding withdrawal of the programme. Their workflow data trains AI agents designed to perform the computer tasks they currently do. 

The executives overseeing the programme hold stock options worth up to $921 million each, tied to a market cap target dependent on AI-driven restructuring. European workers are exempt from the surveillance by law. American workers have no opt-out.

  • Behavioural data exists on every remaining employee. Meta says it won't be used in performance reviews. Who enforces that?
  • Should surveilled staff record their own logs to prove their case when the MCI system glitches and they are disciplined or sacked?
  • Should companies disclose when executives profiting from AI restructuring are covered by different employment protections than the workers generating the training data?
  • Does mandatory, no-opt-out surveillance on a workforce facing mass layoffs constitute interference with the right to organise under the NLRA?

3 responses from the Newspage community

Copy all

Star Quote
Copy

Colette Mason
AI Ethics Consultant at Clever Clogs AI
Fortune reported Meta's SEC filing answers the disclosure question on its own terms: executives hold stock options worth up to $921 million each, triggered by a $9 trillion market cap target that depends on AI replacing human work.

The people generating the training data to reach that target have no opt-out, no equity stake in the outcome, and no GDPR protection, because they're in the US, not Europe. Alexandr Wang, hired from Scale AI for $14.3 billion to run Superintelligence Labs, told Andreessen Horowitz the biggest blocker to building agent capabilities was a lack of data. The employees became the supply chain to their own replacement.

The leaked memo from Janelle Gale, Meta's chief people officer wrote "This is not an easy tradeoff and it will mean letting go of people who have made meaningful contributions to Meta during their time here."

Is it right that a small group get $921m payoffs at the expense thousands of laid-off skilled workers and their impoverished dependents?
Copy

Editor
Steven Greenall
Director at Greenall Estate Planning
Simply big brother on steroids. Who on earth will enjoy working under that scrutiny
Copy

Editor
Paul Denley
CEO at Oakham Wealth Management
The industrial revolution mechanised labour. AI risks mechanising judgement itself. There is something deeply unsettling about workers training systems designed to replace them while every keystroke is harvested without consent or opt out. If surveillance becomes the price of employment, the line between productivity management and digital Taylorism starts to blur.

The contrast with Europe is revealing. GDPR treats behavioural data as worthy of protection; US workers are told to trust assurances the data “won’t be used” against them. But trust without independent oversight is not governance - it is PR.

And this may backfire. If AI defines what “good” work looks like, firms may drive out the unconventional thinkers they should value most. Some of humanity’s greatest breakthroughs were accidental, serendipitous or dreamt up far from a desk. Meta may be training the AI. The bigger question is whether it is training its best people to leave.