Scale AI's 2023 Funding-Fueled Surge
Scale AI
Scale’s 2023 jump showed that AI data vendors had become direct picks and shovels for the foundation model funding boom. As capital flooded into OpenAI, Anthropic, Cohere, Runway, and Harvey, those companies urgently needed humans to rank outputs, fix bad answers, and create reward data for RLHF, which turned Scale from an autonomous vehicle labeling business into a core training supplier for frontier labs and generative AI teams.
-
The mix shift was dramatic. Scale went from about $290M of 2022 revenue tied mainly to autonomous vehicle and defense labeling to roughly $760M ARR in 2023, as autonomous driving demand weakened and LLM training demand replaced it.
-
This was not ordinary SaaS expansion. The revenue spike tracked financing velocity at model labs, because new funding was quickly converted into outsourced human work, rater operations, and evaluation pipelines needed to ship and improve frontier models.
-
The same pattern created a new category around human feedback infrastructure. Invisible grew to an estimated $134M in 2024 on contracts with Microsoft, Cohere, AI21, Mistral, and Perplexity, showing Scale was riding a market wide post training spending wave, not a one off contract cycle.
Going forward, the winning vendors will be the ones that turn one time labeling surges into durable evaluation and compliance workflows. As labs move from broad crowd tasks to expert judgment, safety testing, and audited human review, the market shifts from raw annotation volume to higher value, harder to replace post training infrastructure.