Google Disentangles from Scale AI Post-Meta Deal
TL;DR:
- Google is ending its partnership with Scale AI after Meta acquired a significant stake in Scale.
- Other AI companies are re-evaluating their relationships with Scale AI.
- This shift is benefiting Scale AI's rivals.
Google is dropping its $200 million-per-year engagement with Scale AI after Meta secured a 49% non-voting stake, valued between $14.3 billion and $14.8 billion, prompting strategic withdrawal from competitors.
Meta’s move, which includes hiring Scale’s CEO Alexandr Wang into a senior role, triggered immediate pushback across the AI industry.
Companies like Microsoft, xAI, and OpenAI are reportedly assessing their ties too, fearing that sharing sensitive prototype data with Scale, now backed by a top rival, jeopardizes their competitive edge.
As one Reuters source noted, “companies that compete with Meta…are concerned that…Scale’s workers…providing data‑labeling services” could expose their roadmaps to a rival.
This shift is already boosting Scale’s rivals: Labelbox expects “hundreds of millions” in new business, while Handshake reports demand tripled overnight.
Some AI labs are opting to build in-house labeling teams to regain control over proprietary data.
Scale AI, founded in 2016 and valued at roughly $29 billion post-deal, recorded $870 million in revenue in 2024.
Google’s roughly $150–200 million spend on specialized human annotation services supported projects such as Gemini, its GPT‑competitive model.
Despite the turmoil, Scale insists it will sustain operations and uphold security standards.
This is a turning point for the AI supply chain. As neutrality becomes decisive, data‑labeling firms are fragmenting. Startups with explicit independence and companies bringing labeling in-house are now the focus. The era of a centralized human‑annotation provider appears to be ending.
