Scale AI
Scale AI supplies the data, evaluation, and safety layer behind AI systems. Its test is whether model builders and enterprises will keep outsourcing high-stakes data work as internal capabilities expand.
The Company
Scale AI began as a provider of labeled training data and developed into a broader AI infrastructure company offering data, model evaluation, safety, and agent-focused services. It works across commercial and government applications where model reliability is a production constraint.
Meta's June 2025 investment and the move of Scale co-founder Alexandr Wang to Meta reshaped the company. Scale remains a private business, but the transaction made independence, customer relationships, and governance central questions for its next phase.
Why It Matters
Scale sits at a bottleneck in the AI economy: trustworthy training data and evaluation. Its trajectory is a read-through on whether value accrues to model companies, infrastructure vendors, or the operational layer that makes models usable in the real world.
Valuation History
What Changed
Scale's data-labeling business became a major supplier to autonomous-vehicle and AI developers.
Raised capital at a $13.8B valuation as generative-AI evaluation demand accelerated.
Meta invested a reported $14.3B for a 49% non-voting stake, valuing Scale at $29B.
THE NEXT COMPANY. A CLEARER PICTURE.
Get the next breakdown.
Independent company research and perspectives on private markets, delivered through The Private Ledger on Substack.