Tensions are running at an all-time high inside Meta Platforms. As the tech giant pours a staggering $135 billion into artificial intelligence infrastructure, it has initiated a brutal workforce restructuring to aggressively transition into an AI-first organization.
At the center of this workplace revolution is Andrew “Boz” Bosworth, Meta’s long-tenured and outspoken Chief Technology Officer. Tasked by CEO Mark Zuckerberg to lead the company’s internal overhaul, Bosworth is spearheading a massive flattening of human management and deploying highly controversial worker-tracking software—all to accelerate the rise of autonomous AI “agents.”
The Reality of the Restructuring: Massive Layoffs and Reassignments
Meta’s aggressive pivot toward artificial intelligence has severely disrupted its traditional workforce structure:
-
The Layoffs: Meta recently laid off approximately 8,000 employees globally to aggressively lean out human teams.
-
The Reassignments: Simultaneously, the company shifted over 7,000 existing workers out of legacy groups and directly into AI infrastructure, silicon development, and data center operations.
-
The Structural Philosophy: Mark Zuckerberg defended the downscaling to employees, stating, “If a team used to take 50 or 100 people and now it takes 10, having 50 or 100 people on that team can actually be counterproductive… we need to fix that.”
The Controversy: Mandatory Keystroke and Mouse Surveillance
The internal unrest escalated dramatically following an announcement that Meta would monitor human activity on work devices to build its next-generation corporate automation systems.
Under a newly rebranded initiative called the Agent Transformation Accelerator (ATA) (originally “AI for Work”), Meta has begun installing software on employee computers to record mouse clicks, keystrokes, and real-time screen context across core applications like Gmail, GChat, and internal dashboards.
The Goal: To capture real-world examples of how engineers, developers, and managers navigate menus, execute keyboard shortcuts, and complete everyday computer tasks, creating training data to build autonomous AI agents that can replicate human workflows.
The initiative immediately faced severe employee pushback, becoming the most negatively received internal announcement in Meta’s history, according to tracking data from the workplace app Blind. More than 1,500 employees signed a petition demanding a halt to the computer-use monitoring.
When cornered by anxious employees on Meta’s internal forum asking how they could opt out of the tracking, Bosworth was uncompromising. He directly stated that there is no option to opt out on company-issued devices, advising staff concerned about extreme privacy to avoid opening personal communications on their work machines.
A Future Where “Agents Primarily Do the Work”
Despite previously leading the company’s heavily criticized and financially draining “Metaverse” push via Reality Labs, Bosworth retains absolute confidence from Zuckerberg. His blueprint for the future of Meta relies on entirely stripping away corporate bureaucracy, planning documents, and human manager layers in favor of code:
[ Traditional Tech Corporate Structure ]
Many Human Managers ➔ Lengthy Written Proposals ➔ Large Engineering Teams
│
▼ (Boz's Overhaul)
[ The AI-First Autonomous Framework ]
Few/No Human Managers ➔ Direct Prototype Testing ➔ AI Agents Doing Tasks
In a recent internal memo layout out the future of Meta’s operations, Bosworth made his long-term expectations clear to the remaining staff:
“The vision we are building towards is one where our agents primarily do the work. Our role is to direct, review and help them improve… We’re already seeing some tasks that used to take hours now take minutes, and soon we won’t need to be in the loop on some tasks at all.”
