Meta’s months-old AI unit is a soul-crushing gulag, say the engineers stuck inside it

2026-06-12 · Source: AI News & Artificial Intelligence | TechCrunch · Field: Business & Management — Corporate Strategy & Leadership, Human Resources & Workforce Development, Operations & Process Management · Depth: Fundamental Awareness, short

Summary

Meta's three-month-old Applied AI unit, comprising approximately 6,500 engineers and product managers, is experiencing severe internal morale issues, with employees describing their work as "soul-crushing" and comparing the unit to a "gulag." Many were "drafted" into the team via surprise emails to generate puzzles and coding problems for AI model training. CEO Mark Zuckerberg justified this by stating Meta employees possess "significantly higher" intelligence than third-party contractors. The unit, led by Maher Saba and reporting to CTO Andrew Bosworth, initially had a structure where up to 50 employees reported to one manager. This internal unrest is part of broader company-wide morale problems, including a petition signed by over 1,600 employees protesting a program monitoring clicks and keystrokes for AI training data. Zuckerberg has acknowledged employee distress in an internal memo.

Key takeaway

For Directors of AI/ML or VPs of Engineering considering rapid internal reassignments for AI initiatives, you must prioritize transparent communication and voluntary participation. Ignoring employee sentiment and forcing "drafts" into "soul-crushing" tasks risks widespread morale collapse, internal revolt, and significant talent attrition, ultimately hindering your AI development goals. Ensure your team structures support engagement, not just efficiency, to avoid the negative outcomes seen at Meta.

Key insights

Aggressive, top-down AI development strategies can severely damage internal employee morale and productivity.

Principles

• Forced internal reassignments erode employee trust.
• High manager-to-employee ratios can lead to disengagement.
• AI training data collection methods impact morale.

Method

Internal employees were drafted to generate puzzles and coding problems for AI model training, based on the premise of their superior intelligence over external contractors.

Topics

• Meta Applied AI
• Employee Morale
• AI Workforce Management
• AI Training Data
• Corporate Culture
• Organizational Restructuring

Best for: CTO, Executive, Director of AI/ML, VP of Engineering/Data, Tech Journalist

Related on AIssential

• Meta is rapidly reorganizing its workers’ jobs around AI: ‘Transfers aren’t optional’ — Meta's aggressive AI pivot involves mandatory job reassignments, management restructuring, and employee monitoring, leading to significant internal dissent.
• Meta Employees Absolutely Hate Mark Zuckerberg’s Plan for a Companywide AI Hackathon — Meta's companywide AI hackathon announcement sparked significant internal employee frustration and disbelief.
• Meta’s AI chief scientist leaves with parting shots — Internal dissent at Meta reveals tensions between established AI research and new LLM-focused strategies.
• Meta Layoffs: 'Success Isn't a Given', CEO Warns — Major tech companies are restructuring workforces and increasing AI investment, driven by AI's impact on productivity and organizational design.
• Does your CEO have AI psychosis? Aaron Levie thinks most of them do. — The tension between AI's perceived job replacement and user rejection of forced AI integration is reshaping the tech landscape and hiring.
• Exclusive: Departing Meta staffer posts biting anti-AI video internally amid mass layoffs — AI-driven corporate shifts can provoke employee resistance, including discussions of data poisoning as a protest method.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI News & Artificial Intelligence | TechCrunch.