Competition Over Customer Strategy Playbooks

The lawsuit mattered less as a legal endpoint than as a signal that AI talent marketplaces are now fighting over customer relationships and operating playbooks, not just worker supply. In this market, a customer strategy deck is practical infrastructure. It can show which model lab is buying coding evaluators versus legal experts, how pricing is packaged, what quality checks are promised, and where expansion opportunities sit. That is why a dispute over internal documents fits the shape of competition between Scale, which built a large managed labeling business, and Mercor, which grew by matching labs with pre vetted experts for RLHF work.1 2 3 4 5

• Scale framed the case around trade secrets tied to customer strategy and product packaging, and court records show the matter was later voluntarily dismissed with prejudice and terminated. That combination suggests the case was serious enough to file, but it did not continue into a public merits ruling.1 2 3
• The underlying rivalry is concrete. Scale sells a broader stack for labeling, model training, and deployment, while Mercor is centered on sourcing and managing expert workers for RLHF tasks. Both are competing for the same frontier lab budgets for human evaluation, post training, and specialized data work.4 5 6
• This is a common pattern in the category. As data labeling moved from generic clickwork toward scarce experts in coding, law, medicine, and reasoning, the most valuable asset became a map of which customers need which experts, at what quality bar, with what workflow. That makes sales knowledge and delivery templates strategically sensitive.5 6 7

Going forward, the category should keep shifting from broad labor pools toward tighter control of expert identity, workflow software, and customer specific delivery know how. The winners are likely to look less like job boards and more like managed infrastructure providers, where process, trust, and embedded customer knowledge matter as much as raw contractor count.4 5 6 7