Meta Invests $14.3 Billion in Scale AI as CEO Alexandr Wang Steps Down to Join Facebook Parent

In one of the biggest developments yet in the AI arms race, Meta has taken a nearly 50% stake in Scale AI, a data-labeling powerhouse critical to the development of advanced artificial intelligence models. The deal reportedly totals $14.3 billion and values Scale at a staggering $29 billion—more than double its valuation just a year ago.

The investment marks a strategic shift for Meta as it works to catch up with AI frontrunners like OpenAI, Google DeepMind, and Anthropic. And it comes with a major leadership shakeup: Scale AI co-founder and CEO Alexandr Wang is stepping down to join Meta, where he will play a central role in shaping its most ambitious AI initiative to date—what insiders are calling its “superintelligence” efforts.

Meta’s Bet on Data—and on Wang

Meta confirmed the deal Friday, saying it plans to deepen its collaboration with Scale AI to “produce data for AI models.” Wang, 27, will help spearhead the company’s next-generation AI programs from within Meta, but will stay involved with Scale as a board director. Meanwhile, Scale’s chief strategy officer Jason Droege will take over as interim CEO.

“We will share more about this effort and the great people joining this team in the coming weeks,” Meta said in a statement.

While the companies are closely aligning, Scale AI stressed that it remains an independent company. The capital infusion will be used both to cash out existing investors and shareholders and to fuel future growth, the company said.

Who Is Alexandr Wang?

Wang, a former MIT dropout, founded Scale AI in 2016 at just 19 years old. The company started by automating data labeling for self-driving cars—a tedious but essential task. In the years since, Scale evolved into the go-to infrastructure provider for AI labs in need of massive, high-quality datasets to train large language models (LLMs) and other advanced systems.

By combining machine learning with human annotators, Scale delivers the “fuel” that powers today’s AI. Its services have been used by OpenAI, Microsoft, Meta, and the U.S. Department of Defense. The company has increasingly focused on hiring PhDs, machine learning experts, and senior engineers to meet the growing demand for premium data annotation in the age of frontier models.

Wang quickly rose to prominence in the tech world, known for his sharp mind and aggressive scaling strategy. His shift to Meta signals not only a talent acquisition for the company, but also a deeper commitment to building world-class AI from the ground up.

Why This Deal Matters

This investment gives Meta a powerful asset at a critical moment. While the company has launched its own open-source LLMs—most recently the LLaMA 3 family—its AI tools have lagged behind rivals in terms of performance and adoption. A recent report by SingalFire found that Meta lost 4.3% of its top AI researchers to competitors in 2024, highlighting its struggles to retain elite talent.

Meanwhile, Meta’s social media business—its cash cow—faces stagnation, regulatory heat, and rising competition from platforms like TikTok. AI represents both a necessary evolution and a fresh battleground for relevance and growth.

This deal also raises eyebrows because of its scale and timing: Just a year ago, Scale raised $1 billion from backers including Amazon and Meta, at a $13.8 billion valuation. With this new round, Meta now holds a controlling interest (49%) and Scale’s valuation has more than doubled—reflecting the heightened importance of training data as a strategic asset in the AI ecosystem.

The Hidden Power of Data Labeling

While AI headlines often focus on flashy chatbots or billion-parameter models, none of it works without data—clean, structured, labeled data. That’s where companies like Scale come in. And that’s why Meta’s move is so telling: it’s not just buying data services, it’s buying into the infrastructure of intelligence itself.

Scale isn’t alone in this space—competitors like Snorkel AI, Labelbox, and Surge AI are also ramping up—but its head start, technical sophistication, and client base put it in a class of its own.

Now, with Wang inside Meta’s AI labs and a near-$30 billion valuation, the line between Big Tech and the startups that power it has never been blurrier.