Who owns Scale AI? Ownership structure explained (2026)

Scale AI sits at the center of one of the most important supply chains in technology: the pipeline that turns raw data into usable training sets for AI models. Every major AI lab — from OpenAI to Google to Meta — needs labeled, structured data to build and refine their systems. Scale built the infrastructure to deliver that at scale, making it a critical vendor in the AI arms race.

That position makes Scale AI's ownership structure unusually consequential. When your largest shareholder is also one of your biggest customers — and your founder leaves to work for that same company — the question of who controls the business carries real strategic weight. The $14.3 billion Meta investment in June 2025 reshaped Scale's ownership overnight, triggering customer defections and raising industry concerns about neutrality.

This article breaks down who owns Scale AI, how that ownership evolved, who holds actual control, and why it matters for the broader AI ecosystem.

Scale AI was founded in 2016 by Alexandr Wang and Lucy Guo, both of whom were teenagers at the time. The company launched out of Y Combinator and is headquartered in San Francisco, California.

The core business is data labeling and data engineering for AI. Scale provides annotation services — tagging images, text, audio, and video — that machine learning models need for training. Over time, it expanded into data management, model evaluation, and AI-powered decision support for both commercial and government clients.

Scale's customer base spans the AI industry and the U.S. defense sector. On May 6, 2026, the Pentagon awarded Scale AI a $500 million contract for data analysis and decision-making support — a fivefold increase from its previous defense deal. The company is valued at $29 billion following its most recent funding round in June 2025, making it one of the most valuable private AI companies in the world.

Scale AI is a private company, so ownership details are less transparent than those of a publicly traded firm. Still, the broad contours are clear: a single massive corporate backer, a roster of elite venture capital firms, and a founder who retains board influence despite stepping away from day-to-day operations.

The most significant ownership event in Scale AI's history occurred on June 12, 2025, when Meta invested $14.3 billion to acquire a 49% economic stake. That single transaction valued Scale at $29 billion and made Meta the company's dominant shareholder by a wide margin.

But economic ownership and control are different things. Reports indicate that Meta holds no voting power in Scale AI. The exact mechanism — whether through a dual-class share structure, specific voting agreements, or another arrangement — is not publicly disclosed. The practical effect is that Meta has a massive financial interest in Scale's success but cannot unilaterally direct the company's strategy or governance.

This structure is unusual. A 49% stake without voting rights suggests a deliberate effort to preserve Scale's independence, at least on paper. For Meta, the investment likely secures priority access to Scale's data labeling capabilities for training its AI models. For Scale, it provided a massive capital infusion while theoretically maintaining operational autonomy.

Before the Meta deal, Scale AI raised capital across multiple rounds from a who's-who of Silicon Valley investors and corporate backers.

The diversity of corporate backers is notable. Nvidia, Amazon, Intel, AMD, and Cisco all participated in the Series F — a sign that Scale's data infrastructure was seen as strategically important across the AI hardware and cloud ecosystem, not just by one player.

Exact ownership percentages for these investors are not publicly disclosed, though Meta's 49% stake necessarily diluted all prior shareholders significantly.

Alexandr Wang remains on Scale AI's board of directors. His exact equity stake is not publicly disclosed, though his net worth was estimated at $3.2 billion in May 2026, suggesting he retains a meaningful position.

Lucy Guo, who co-founded the company but departed in 2018, holds a reported 3% stake.

No significant foreign ownership subject to regulatory scrutiny has been identified.

The gap between economic ownership and operational control is the defining feature of Scale AI's governance — and the most important thing to understand about who actually runs the company.

Jason Droege assumed the role of Interim CEO in June 2025, following Alexandr Wang's departure. Droege is not a founder. He previously founded Uber Eats and held leadership roles at Axon before joining Scale AI as Chief Strategy Officer in September 2024.

His appointment was part of the broader leadership transition triggered by the Meta investment. Whether Droege will remain in the role permanently or a new CEO search is underway has not been publicly confirmed.

Wang co-founded Scale AI in 2016 and served as CEO until June 2025, when he left to become Chief AI Officer at Meta — the same company that had just acquired 49% of Scale. He remains on Scale's board of directors, maintaining formal governance influence.

The optics are striking: the founder of an AI data company that serves multiple competing AI labs now works for the company that owns nearly half of it. This dynamic is central to the independence concerns discussed below.

The identity of Scale AI's board chair is not publicly disclosed. What is known is that Meta holds no voting power despite its 49% economic stake. This strongly implies a multi-class share structure or specific contractual voting arrangements, though the exact details remain private.

No single individual is confirmed to hold more than 10% of voting power. The practical effect is that control likely rests with a combination of Wang (via board presence and potential supervoting shares), early investors like Founders Fund and Accel, and the management team.

Scale AI's ownership has evolved rapidly, with its valuation jumping from startup to $29 billion in under a decade.

The trajectory is clear: each round brought larger checks, higher valuations, and deeper ties to corporate AI players. The Series G was a step change — not just in dollars but in the fundamental ownership dynamic of the company.

The most significant controversy surrounding Scale AI's ownership is the fallout from Meta's 49% investment. Following the June 2025 deal, several major AI companies — including OpenAI, Google, and xAI — reportedly cut or reduced their work with Scale AI.

The concern is straightforward: if Meta owns half the company and the founder now works at Meta, can Scale be trusted as a neutral vendor? AI labs that compete directly with Meta have reason to worry that their proprietary data labeling workflows, model performance insights, or strategic priorities could leak — even unintentionally — to a competitor.

OpenAI publicly stated its decision was unrelated to the Meta deal. Scale's management has maintained that Meta holds no voting power and the company remains independent. But perception matters in enterprise sales, and the customer concentration risk is real.

In September 2025, Scale AI filed a federal lawsuit against rival AI firm Mercor, alleging that a former Scale employee downloaded over 100 confidential documents and shared them with Mercor while still employed at Scale. The case remains ongoing and underscores the competitive intensity in the AI data infrastructure market.

The departure of both the CEO (Wang, June 2025) and the CFO (Cinelli, January 2026) within a seven-month window is notable. Losing both the founder-CEO and the finance chief in quick succession creates execution risk, particularly for a company navigating a complex ownership transition and customer relationship challenges simultaneously.

Scale AI's ownership structure is not just a financial curiosity — it has direct implications for the AI industry's competitive dynamics.

When a company that provides data labeling services to multiple competing AI labs is nearly half-owned by one of those labs, neutrality becomes a strategic question, not just a governance one. The customer defections following the Meta deal illustrate this clearly: ownership shaped purchasing decisions across the industry.

For Scale's government clients, including the U.S. Department of Defense, the independence question carries national security dimensions. The $500 million Pentagon contract suggests confidence in Scale's operational autonomy, but the governance structure will likely face ongoing scrutiny as AI becomes more central to defense applications.

Ownership also determines Scale's strategic direction. Whether the company pursues an IPO, deepens its Meta relationship, or pivots toward government work will depend on who holds actual decision-making power — a question that remains partially opaque.

Jason Droege serves as Interim CEO of Scale AI. He assumed the role in June 2025 after co-founder Alexandr Wang departed to become Chief AI Officer at Meta. Droege previously founded Uber Eats and held leadership positions at Axon before joining Scale as Chief Strategy Officer in September 2024.

No. Scale AI is a privately held company. It has not filed for an IPO. The company was most recently valued at $29 billion following its June 2025 Series G funding round.

Scale AI was co-founded in 2016 by Alexandr Wang and Lucy Guo. Wang served as CEO until June 2025 and remains on the board. Guo departed the company in 2018 and holds a reported 3% equity stake.

Meta is the largest shareholder, holding a 49% economic stake acquired in June 2025 for $14.3 billion. Other significant investors include Accel, Founders Fund, Tiger Global Management, Greenoaks Capital, Nvidia, and Amazon. Exact ownership percentages for investors other than Meta are not publicly disclosed.

Despite owning 49% of Scale AI's economic interest, Meta reportedly holds no voting power. The exact governance mechanism — likely a multi-class share structure or specific voting agreements — is not publicly disclosed. Scale's management has publicly maintained that the company operates independently.