Meta’s $14.3B Stake Triggers Trust Crisis, Major Clients Exit Scale AI

Meta’s massive $14.3 billion investment in data-labeling unicorn Scale AI is sparking an unexpected client exodus. Following Meta’s acquisition of a 49% stake, key customers like Google, Microsoft, and OpenAI are swiftly scaling back or exiting their partnerships, citing growing concerns over data sovereignty and competitive exposure.

According to Reuters, Google had planned to pay Scale up to $200 million this year for data services but is now shifting its business to competitors. Microsoft is also reportedly reducing its usage, while OpenAI had already begun diversifying suppliers months ago, despite publicly maintaining Scale as one of its vendors.

At the heart of this fallout is a rupture in data neutrality. As Scale aligns closely with Meta, clients fear that proprietary data and product prototypes could be indirectly exposed to a major AI competitor. A Google insider stated that continuing the partnership is akin to “leaking our roadmap to a rival.”

The fallout is benefiting rivals. Labelbox anticipates hundreds of millions in new contracts, while Handshake claims client inquiries have tripled. The Meta deal appears to have shattered the industry’s expectation of neutrality in data infrastructure.

Further fueling the disruption, Meta has hired Scale’s 28-year-old CEO Alexandr Wang to lead its new “Super Intelligence” lab, reporting directly to Mark Zuckerberg. Analysts see this as a sign of Meta’s deep concerns over falling behind in AI; internally, its Llama model is seen as lagging the market, prompting Meta to license rival models like Claude for use in its Devmate coding assistant.

While Scale insists it will continue to operate independently, the trust gap may be irreversible. As data becomes a strategic asset, this capital-driven rift could reshape the foundational layer of the AI ecosystem.