14.06.2025 06:56

Google, a major client of data labeling company Scale AI, is reportedly pulling out of a significant contract, according to internet sources. This decision follows Meta's recent acquisition of a 49% stake in Scale AI, dramatically increasing the startup's valuation from approximately $14 billion to $29 billion. The shift is substantial, considering Google's projected expenditure of around $200 million this year on Scale AI's services for training its Gemini AI model.

Already this week, Google has initiated discussions with alternative providers, intending to significantly reduce its reliance on Scale AI for data labeling. This strategic move represents a considerable blow to Scale AI, which had anticipated substantial revenue from Google’s contract. Despite the setback, however, Scale AI intends to continue operations. A planned transition will see CEO Alexandr Wang and some employees join Meta, while the majority of the workforce remains with Scale AI.

While Scale AI assures its relationships with major corporations and government bodies remain strong and that data security remains a priority, they refrained from commenting specifically on Google's decision. The company’s 2024 revenue totalled $870 million, with Google contributing an estimated $150 million last year. This loss, however, is part of a broader trend among tech giants.

The trend towards reducing reliance on Scale AI extends beyond Google. Reports suggest Microsoft is reevaluating its engagement with the company, while xAI, Elon Musk's venture, plans to terminate its contract. Even OpenAI, despite continuing to use Scale's services, has scaled back its usage in recent months. These actions reflect a growing concern amongst AI developers regarding potential data exposure.

This apprehension stems from the nature of the data labeling process, which often involves sharing prototype tools and sensitive datasets. With Meta holding a near-majority stake in Scale AI, competitors fear the risk of their proprietary research and technological details being inadvertently disclosed to a direct rival, potentially compromising their competitive advantage in the rapidly evolving AI landscape.