Meta Platforms (META) has agreed to acquire a 49% stake in artificial intelligence startup Scale AI for $14.8 billion, according to The Information, citing two people familiar with the matter. While the deal is not yet finalized, it marks a significant step in Meta’s push to expand its AI capabilities.
Founded in 2016, Scale AI (SCAI.PVT) specializes in providing large volumes of labeled and curated training data—critical for building advanced AI models such as OpenAI’s ChatGPT. The reported deal structure could benefit early backers of Scale AI, including Accel, Index Ventures, Founders Fund, and Greenoaks Capital, along with current and former employees.
As part of the agreement, Scale AI CEO Alexandr Wang is expected to take on a top role at Meta, leading a newly created “superintelligence” lab. The move aligns with CEO Mark Zuckerberg’s broader efforts to ramp up Meta’s AI initiatives through aggressive talent acquisition and product development.
Meta has faced skepticism in recent months over its position in the AI race, especially after its initial Llama 4 language models underperformed expectations. The company also delayed the release of its flagship “Behemoth” model over capability concerns, The Wall Street Journal reported last month.
To avoid potential antitrust scrutiny—already a concern for Meta due to past acquisitions like Instagram and WhatsApp—the deal’s structure may be crafted to reduce regulatory risk, according to The Information.
Scale AI was last valued at $13.8 billion during a funding round in spring 2024. The company generated approximately $870 million in revenue last year and expects to surpass $2 billion in 2025. Its client base includes major AI players like OpenAI and Cohere, as well as tech giants Microsoft, Meta, and Cisco Systems.
Meta’s deepening collaboration with Scale AI aligns with its aggressive push into artificial intelligence, as the company prepares to compete head-on with other Silicon Valley giants in the race toward artificial general intelligence (AGI). Meta plans to invest up to $72 billion in AI initiatives over the next year, ramping up its chip inventory, expanding data center infrastructure, and accelerating AI product development. “This is going to be intense,” CEO Mark Zuckerberg said in January, calling 2025 a defining year to “set the course for the future.”
Meanwhile, Alexandr Wang, Scale AI’s CEO, has emerged as a key voice in Washington’s AI policy debates. He has publicly warned about the rise of Chinese AI startup DeepSeek and has directly appealed to the Trump administration to increase federal investment in AI. In April, Wang testified before Congress, advocating for the creation of a national AI data reserve to help maintain U.S. leadership in the field.
Wang’s federal focus is deeply personal—he grew up in Los Alamos, New Mexico, the son of two weapons physicists. After early roles at Addepar and Quora, Wang left MIT to co-found Scale AI in 2016 with Lucy Guo, a fellow Quora alum who departed the company in 2018.
Inside Scale AI’s Operations
Scale AI employs a global network of human contractors to annotate text, images, and video, creating the high-quality training data essential for fine-tuning AI systems. The company originally targeted the autonomous vehicle sector before shifting focus to generative AI.
Operating across roughly 9,000 locations worldwide, Scale AI has faced scrutiny over its labor practices, including allegations of wage theft and worker misclassification. Earlier this year, the U.S. Department of Labor concluded a nearly year-long investigation into the company’s employment practices.
Meta, Scale AI, and the involved investors declined to comment, Reuters reported.
Pump Radar
Stock Alerts