Meta’s $15B power move to poach Scale’s CEO

Meta just dropped what could be its biggest AI power play yet—a massive ≈$14–15 billion investment for roughly 49 % of data-labeling heavyweight Scale AI, valuing it around $28–29 billion.

In the past week, reports surfaced that Meta is preparing a monumental deal to grab half ownership in Scale AI—firmly anchoring itself in the data annotation engine that underpins giants like OpenAI, Google, and Microsoft

Meta’s latest Llama 4 model flopped on reasoning tasks, and its “Behemoth” AGI effort hit turbulence. Enter Scale—known for its meticulous labeling and annotation prowess, plus its gig-scale workforce powering AI pipelines .

As part of the merger-of-minds, Scale CEO & cofounder Alexandr Wang (just 28!) will join Meta’s newly minted “superintelligence lab”, leading from the front while Scale retains independence—Jason Droege steps in as interim CEO.

But there are red flags. Scale’s major clients—Google, Microsoft, xAI—are reportedly pulling their data projects, spooked by the conflict of interest. Google alone pays around $200 million annually to Scale.

So, what’s the bottom line?

• Data control: Meta locks down access to the high-quality labeled data crucial for advanced AI training.
  
• Superintelligence push: With Wang onboard, Meta seals up talent while building its AGI team of ~50 researchers under Zuckerberg’s leadership.
  
• Competitive cliffhanger: Clients fleeing Scale could gut the startup’s model, but Meta’s financial heft pledges stability.
  
• Labor spotlight: Scale’s gig workforce—Remotasks—faces criticism for poor wages and conditions, raising questions about ethics in this mega-dea.l

🔑 Key Points

• $14.3–15 B for 49 % of Scale AI → values Scale at $28–29 B
  
• Alexandr Wang joins Meta’s AGI lab, Scale retains independence
  
• Major clients exiting (Google, Microsoft, xAI) amid conflict concerns
  
• Meta pushing superintelligence, building a 50‑person elite team
  

Ethics under microscope: Worker welfare in data labeling spotlighted