Meta's three-month-old Applied AI team, a 6,500-person unit of engineers and product managers tasked with generating training data for the company's AI models, is in open revolt. An employee-only livestream this week was hijacked by an expletive-laden tirade demanding attendees tell a senior AI executive he was "a piece of sh_t," according to a Wired report. One presenter covered their face with their hands. Meta CEO Mark Zuckerberg addressed the unrest in an internal memo Friday, acknowledging recent changes had "caused distress."
The grievance is structural. Employees describe being assigned to the unit with no real choice — join or quit — and refer to themselves as "draftees." Many learned of their reassignment through a surprise email, a process one engineer described on Reddit as "quite random." Their new job is generating puzzles and coding problems used to train Meta's AI models, work several employees told Wired they find "soul-crushing."
More than 1,600 employees company-wide have separately signed a petition protesting a monitoring program that records clicks and keystrokes to produce AI training data. Chris Cox, Meta's chief product officer, felt compelled to address what he called the "brutal" environment on a call with employees this week.
Key facts
- 01Meta's three-month-old Applied AI team consists of roughly 6,500 engineers and product managers, many drafted via surprise email.
- 02More than 1,600 Meta employees signed a petition protesting a program monitoring their clicks and keystrokes for AI training data.
- 03The unit reports to Maher Saba, a 12-year Meta veteran from Reality Labs, the division that spent $83 billion on the metaverse.
- 04Alexandr Wang sold Scale AI to Meta for $14.3 billion before taking the chief AI officer role at Meta Superintelligence Labs.
- 05Up to 50 employees originally reported to a single manager inside the new organization.
The logic behind drafting internal engineers rather than hiring outside contractors came directly from Zuckerberg. In a leaked audio recording from an April internal meeting, he argued that Meta's average employee has "significantly higher" intelligence than third-party labelers, and that Alexandr Wang — who sold Scale AI to Meta for $14.3 billion before becoming chief AI officer and heading Meta Superintelligence Labs — knows the data-labeling business well enough to know the difference matters.
That $14.3 billion price tag for Scale AI sets the financial frame. Meta is treating high-quality training data as a strategic asset worth paying enormous sums for, and the Applied AI team is the in-house extension of that bet. An April internal announcement reviewed by Business Insider acknowledged that Meta's models still lacked the knowledge to outperform humans on technical tasks like coding, and that closing the gap required new training examples.
Hence the draft. The internal memo framed it directly:
“For agents to understand how people actually complete everyday tasks using computers, we need to train our models on real examples”— Meta, Internal announcement, April 2026
The Applied AI team is led by Maher Saba, a 12-year Meta veteran who previously served as a vice president in Reality Labs — the division that burned through $83 billion building the metaverse before the company pivoted its capital and narrative to AI. Saba's organization reports up to Meta CTO Andrew Bosworth. At launch, the unit was structured so that up to 50 employees reported to a single manager, a span of control that helps explain why so many draftees describe the environment as impersonal.
None of this exists in isolation. Meta has executed repeated rounds of layoffs over the past several years, accelerating as AI capex climbed, and the cultural cost is now showing up in places leadership cannot ignore. A hijacked livestream is not a typical engineering-org problem. Neither is a 1,600-signature petition against employee surveillance.
Zuckerberg's Friday memo conceded mistakes and pledged fixes, closing with a familiar line:
The deeper issue for Meta is whether the Applied AI bet survives its own internal politics. Drafting 6,500 of your highest-paid engineers to write training puzzles is a defensible decision when the alternative is paying Scale-style contractors at lower quality — but only if those engineers stay. If the best ones leave for OpenAI, Anthropic, or Wang's old competitors at the first vesting cliff, Meta will have paid $14.3 billion for a data pipeline and lost the talent it was meant to feed. The next signal to watch is not another memo. It is the attrition number in Applied AI's first full quarter.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
