Surge AI
| Company type | Private |
|---|---|
| Industry | Data annotation, Artificial intelligence |
| Founded | 2020 |
| Founder | Edwin Chen |
| Headquarters | San Francisco , United States |
Area served | Worldwide |
Key people | Edwin Chen (Founder) |
| Products | Data annotation services, RLHF, RL environments, language data annotation |
Number of employees | 1000000 (annotators) (2025) |
| Website | dataannotation |
'Surge Labs Inc., known commercially as Surge AI, is a multinational American enterprise headquartered in San Francisco that specializes in data annotation services. The company focuses on reinforcement learning from human feedback (RLHF), creating RL environments, and annotating language data. Its clientele features prominent organizations such as OpenAI, Google, Microsoft, Meta, and Anthropic. Surge AI ranks among the leading global firms in data labeling.
History
Established in 2020 by Edwin Chen, Surge AI was born out of his dissatisfaction with the quality of crowdsourced data labeling. By mid-2025, the company continued to operate without external investment, remaining entirely bootstrapped by Chen.
Surge initially concentrated on areas like search technology, recommendation systems, and content moderation. It developed a platform designed to match annotators with tasks aligned to their expertise. These annotators contribute to enhancing generative AI models—including chatbots and text-to-image applications—by evaluating model outputs and providing specialized feedback.
As of 2025, Surge AI employed approximately one million annotators worldwide. Although it had not secured outside funding, discussions commenced in July 2025 with investors such as Andreessen Horowitz, Warburg Pincus, and TPG Inc. Surge’s market valuation was estimated to fall between $15 billion and $25 billion.
Controversies
In 2023, reports from The Verge and New York Magazine revealed that Surge AI managed several platforms including Get Hybrid, Task Up, and Data Annotation. The Data Annotation platform attracted criticism for its lack of clarity regarding ownership and unexplained deactivation of annotator accounts. In May 2025, Surge AI faced a class action lawsuit accusing the company of intentionally misclassifying data annotators as independent contractors to avoid granting employee benefits and improperly withholding wages.
In July 2025, two confidential internal documents from Surge AI were leaked: one outlining procedures for model training and AI safety protocols, and another specifying permitted and prohibited websites for contractors involved in training Anthropic’s models.