Meta Platforms Inc. (NASDAQ:META) forcibly reassigned at least 1,000 of its top engineers to a new artificial intelligence division in April, a move employees have dubbed "the draft," which seeks to accelerate AI development at the cost of widespread internal dissent.
"Meta was no longer seeing us as partners," one technical employee told reporters, a sentiment that captures the growing disconnect between management and staff inside the $1.2 trillion technology giant.
The reorganization is part of a massive strategic pivot that includes plans for 8,000 layoffs, representing about 10 percent of the workforce, even as the company increases its full-year 2026 capital expenditure forecast to between $125 billion and $145 billion. The spending is largely directed at securing the data centers and chips needed to compete with rivals like Google, OpenAI, and Anthropic. For comparison, Q1 capex was $19.8 billion, up 45 percent year-over-year.
For investors, the draft is the sequel to Meta's "Year of Efficiency," a period of ruthless cost-cutting that drove significant margin expansion and shareholder returns. Now, the company is risking that discipline and its own cultural fabric on a high-stakes bet that it can build the next generation of AI, even if it has to conscript its own workforce to do it.
The Draft and the Damage
The internal atmosphere at Meta has been described as "historically low," according to reporting from WIRED. The forcible transfers to the new Applied AI Engineering division, which builds tools for research scientists, came with an ultimatum: move or be laid off. This is an unusual tactic in Silicon Valley, where technical staff typically have internal mobility during restructurings.
Morale has been further eroded by widening pay gaps and job insecurity. While thousands face layoffs, CEO Mark Zuckerberg has reportedly offered top AI researchers compensation packages worth as much as $100 million a year. Meanwhile, median total compensation for employees fell to $388,200 last year from $417,400 in 2024, partly due to a 5 percent cut in the stock portion of annual raises. One UK-based group of Meta employees has begun a formal union drive with United Tech and Allied Workers in response to what organizers call "cruel and short-sighted behaviours."
A Surveillance State to Train the Successor
Compounding the anxiety is the mandatory installation of mouse-tracking software, known internally as the Model Capability Initiative (MCI), on the corporate laptops of U.S. employees. The tool captures clicks and navigation patterns to generate data for training AI models that can automate tasks. According to Meta spokesperson Tracy Clayton, "safeguards are in place to protect sensitive content, and the data is not used for any other purpose."
Employees have pushed back, distributing anonymous flyers at multiple U.S. offices that invoke the National Labor Relations Act (NLRA), which protects workers' rights to organize for better conditions. The flyers, seen by Reuters, ask, "Don't want to work at the Employee Data Extraction Factory?" The NLRB has been explicit that using AI for employee surveillance in a way that could interfere with organizing is illegal, placing Meta's program in a legally sensitive position.
The Price of a 'Code Red'
Zuckerberg's move mirrors similar "all-in" moments in tech history. It draws a direct parallel to Google's late-2022 "Code Red," when founders Larry Page and Sergey Brin were pulled back in to ship a competitor to ChatGPT. That effort produced Gemini but also led to an exodus of senior AI talent to rivals. Similarly, Microsoft's early Azure reorganization and Apple's original iPhone project involved conscripting engineers, delivering products but also bleeding talent within 18 to 36 months.
The current situation at Meta appears to be following this pattern. According to WIRED, employees who can afford to leave are hoping to be laid off to receive a severance package that includes a minimum of 16 weeks of pay. This indicates a fundamental breach of the psychological contract between the company and its most valuable employees.
For investors who have cheered Meta's efficiency, the current strategy presents a new kind of risk. The company is spending whatever it takes to secure top AI talent and compute power, a move that has rewarded the stock price. However, the internal backlash, potential for talent drain to competitors like xAI and Mistral, and legal challenges from labor organizing create significant execution risk. Meta shares are down about 5 percent this year, and the success of this AI pivot will depend on whether Zuckerberg can build the future without losing the people he needs to get there.
This article is for informational purposes only and does not constitute investment advice.
