April 29, 2026
Production With a Pulse: Meta is Using its Employees to Train Their Own AI Replacements
The psychological contract between employer and employee used to be implicit. You give your time, skills, energy, and loyalty. In return, you receive fair compensation, reasonable security, and the basic dignity of being treated as a person. For decades now, it has felt as if that contract was being unravelled slowly – though somewhat carefully – as employers continued to ask more of their people in exchange for a lot less. Meta, however, just blew any notion of subtlety or humanity out of the water.
The tech giant recently announced it would be installing tracking software on its US employees’ work computers to capture mouse movements, clicks, keystrokes, and screenshots. This data will be feeding directly into their Model Capability Initiative (MCI), so that the AI could learn from real examples of how people actually use computers. This includes things like mouse movements, clicking buttons, using shortcuts, and navigating dropdown menus on sites such as Google, LinkedIn, Slack and GitHub.
To add insult to injury, a few days later, Meta’s head of HR circulated a memo, confirming that approximately 8,000 employees – roughly 10% of their global workforce – would be made redundant by 20 May 2026, with additional cuts planned for the second half of the year.
The layoffs are apparently structural rather than performance-based. Meta is reorganising teams into AI-focused ‘pods’, creating new role categories like ‘AI builder’, ‘AI pod lead’ and ‘AI org lead’. The company’s internal language describes the goal as “fundamentally rewiring how we operate.”
Note the use of the word ‘rewiring’ – not ‘restructuring’ or ‘reorganising’. Rewiring. The language of circuitry applied to human beings.
Dataset with a Day Pass
Meta is trying its best to frame this as innovation. In reality, it’s nothing more than extraction. It goes way beyond the surveillance or productivity paranoia that we have seen proliferating since Covid lockdowns and the formal dawn of remote and hybrid working. People are literally being asked to train the very technology that will make them all obsolete.
This didn’t just happen, though. The warning signs have been there since at least 2024 when the person now leading Meta’s Superintelligence Labs, Alexandr Wang, formerly of Scale AI – a company built on harvesting workflow data from contractors – stated that “For a lot of the capabilities we want to build into the models, the biggest blocker is actually a lack of data. There’s no pool of really valuable agent data just sitting around anywhere. So we have to figure out how to produce really high quality data.”
Meta’s US employees are that data. They neither agreed, nor volunteered to be, but CTO, Andrew Bosworth has made it clear to all employees that “there is no option to opt out of this on your work-provided laptop.”
Employees in Europe are spared this fate for now. This is thanks mostly to the strict data protection and privacy laws contained in the GDPR.
Meta is not a cautionary tale from some parallel universe. Instead, it is the leading edge of a wave moving through every sector, at different speeds, with different degrees of transparency about what is actually happening. In internal Meta communications, Bosworth has described a future in which AI agents “primarily do the work” while employees “direct, review and help them improve.” It’s therefore safe to assume that it’s the most honest version of what Meta’s MCI is really building toward. That means that the surveillance and the ‘rewiring’ are not separate projects.
And yet, even as Meta mines its workforce for data, it is also building an AI version of its CEO to simulate connection with that same workforce. The surveillance story and the avatar story are not separate; they are the same story told from opposite ends.
The Existential Threat
There is a growing body of evidence showing that the way organisations treat employees has a direct and measurable impact on consumer trust, brand loyalty, and ultimately, revenue. Research indicates that nearly a third of consumer identify employee treatment as a decisive factor in their loyalty to a brand, and a significant proportion have stated that their willingness to try a new company is influenced by what they learn about how that organisation treats its workforce.
Consumers increasingly view employee experience as a proxy for corporate values, drawing conclusions about integrity, responsibility, and ethics based on workplace practices that once remained largely invisible to the public.
Academic studies reinforce this connection, demonstrating that when consumers learn that employees are treated poorly, their perception of the brand declines and their likelihood of purchasing decreases, even when the product itself remains unchanged. This shift in perception doesn’t require scandal-level exposure either. The mere suggestion that employees are undervalued or exploited can be enough to erode goodwill that took years to build.
As organisations invest heavily in automation and AI, public awareness of workforce treatment continues to grow, fuelled by social media, investigative journalism, and an increasingly values-driven consumer base. Customers are paying attention not only to what companies produce, but how those companies operate. And the distance between internal decisions and external reputation has grown remarkably short.
This convergence of employee experience and consumer expectation places leaders at a critical crossroads. The organisations that succeed in the coming decade will be those that recognise the interdependence between technological advancement and human trust, understanding that innovation without confidence in the motives rarely delivers sustainable advantage.
Time to Meet the Moment
Production has always relied on people. Even in the most automated environments, human judgment, creativity, and accountability remain the foundations on which meaningful progress is built. The challenge facing leaders now is how to pursue AI without reducing people to a temporary stepping stone in the pursuit of profits and efficiency.
