Surge AI company profile: valuation, funding & revenue

This Surge AI company profile covers the company's valuation, revenue, funding rounds, investors and history. Surge AI works in artificial intelligence, with a focus on human data for AI training (RLHF). It was founded in 2020 and is headquartered in San Francisco, California, United States. Every figure links to a public source, and anything the company hasn't disclosed is marked as such.

Revenue

About US$1.2 billion — Annual revenue, 2024 (reported by Bloomberg; company said 'well north of US$1 billion'), as of 2024. Source: Wikipedia (citing Bloomberg).

Funding history

No funding rounds recorded from reliable sources.

Surge AI timeline

Key dates from the sources on this page, oldest first. Metric snapshots are listed after the funding events.

• 2020 — Surge AI is founded in San Francisco by Edwin Chen (CEO).
• 2025 — Employees: About 110.
• July 2025 — Reported fundraising talks: Talks with investors including Andreessen Horowitz, Warburg Pincus and TPG reported in 2025, with valuations discussed between US$15 billion and US$25 billion.
• September 2025 — Founder stake: Edwin Chen reported to own about 75%.

Investors

Overview

Surge AI provides the high-quality human data that AI companies use to train and improve their models. Its workforce of skilled annotators writes examples, ranks model responses, builds evaluation tests and creates training environments, particularly for reinforcement learning from human feedback (RLHF), the technique that made chatbots such as ChatGPT and Claude helpful and safe to use. Its customers have included Anthropic, OpenAI, Google, Microsoft and Meta, as well as the US Army.

Edwin Chen founded Surge AI in San Francisco in 2020 and is its CEO and majority owner. Chen previously worked on machine learning and data at Google, Facebook and Twitter, where he became frustrated with the poor quality of crowdsourced data labelling. He built Surge to focus on quality rather than volume. What makes Surge unusual is that it grew to more than a billion dollars in annual revenue without raising any venture capital.

A bootstrapped giant

Most AI companies raise enormous sums from investors. Surge took no outside funding for years, funding its growth from revenue, with a very small team of around 110 employees in 2025. Its spokesperson said revenue in 2024 was "well north of US$1 billion", and Bloomberg reported about US$1.2 billion. That makes Surge one of the most capital-efficient technology companies in history, with revenue per employee far above typical software companies, though much of that revenue is paid out to its large network of contractors.

How Surge makes money

AI labs pay Surge for data projects: for example, thousands of carefully written coding problems with expert solutions, rankings of model answers on difficult medical questions, or adversarial prompts to test a model's safety. Surge recruits and manages annotators, many with specialised skills, through its platform and its DataAnnotation brand, and charges labs for the finished work. Quality is its selling point; it says it uses rigorous testing and monitoring to ensure annotators produce accurate, thoughtful data.

Funding and valuation

As of mid-2025, Surge had not raised outside capital. Reports that summer said it was in talks with investors including Andreessen Horowitz, Warburg Pincus and TPG about its first external raise, with valuations discussed between about US$15 billion and US$25 billion. We haven't seen confirmation that a round closed or at what valuation, so we show valuation as not disclosed. Chen was reported to own about 75% of the company, which would make him one of the wealthiest founders in AI on paper.

Controversies

Surge has faced criticism and legal challenges. Workers on its DataAnnotation platform have complained about unclear ownership of the platform and unexplained account deactivations. A 2025 class action lawsuit alleged that annotators were misclassified as independent contractors rather than employees, denying them benefits. Also in 2025, internal documents, including training guidelines and a list of websites used or excluded for training one client's models, were leaked. These issues highlight the labour and confidentiality risks in the AI data business.

Competition

Surge competes with Scale AI, Mercor, Turing, Invisible Technologies and other data providers, as well as AI labs' own data teams. After Meta's large investment in Scale AI in 2025, several labs reportedly shifted work to independent suppliers such as Surge, which benefited from being founder-owned with no strategic investor.

Surge AI at a glance

Surge AI is a San Francisco AI data company founded in 2020 by Edwin Chen, who owns about 75%. It reported about US$1.2 billion in revenue for 2024 with around 110 employees, without venture funding. Reports in 2025 described talks for a first outside raise at a valuation of US$15 billion to US$25 billion. Its customers have included Anthropic, OpenAI, Google, Microsoft and Meta.

What RLHF is and why it matters

Reinforcement learning from human feedback trains a model by showing it many examples of good and bad responses as judged by people. Annotators compare two answers and say which is better, or rate answers against guidelines, and the model learns to produce responses people prefer. RLHF was key to turning raw language models into useful assistants. As models improve, the feedback required becomes more specialised: judging an expert-level answer in law or mathematics requires a genuine expert. That's why demand has shifted from large volumes of simple labels to smaller volumes of expert judgement, which is more valuable per task.

The economics of human data

Human data companies sit between AI labs, which have enormous budgets, and a global workforce of contractors. Their margins depend on pricing, contractor pay rates and efficiency. Large customers can negotiate hard, and labs may build in-house teams or use synthetic data to reduce reliance on outside suppliers. On the other hand, quality is hard to guarantee, and labs are often willing to pay a premium for suppliers they trust. Surge's ability to stay profitable without outside capital suggests strong pricing power, at least for now.

Questions about AI data suppliers

• How does the supplier test and monitor annotator quality?
• How is client data and intellectual property protected?
• How are contractors classified and paid, and what legal risks follow?
• How dependent is the supplier on a few large customers?

Why quality is so hard in data labelling

Early data labelling businesses relied on huge crowds of low-paid workers completing simple tasks, such as drawing boxes around cars in photos. That approach struggles with the work AI labs now need, because a wrong judgement on a subtle legal question or a piece of code can teach a model the wrong lesson. Surge's founder has argued that the industry underinvested in quality for years, focusing on volume and price. Surge screens annotators with demanding tests, monitors their work continuously and removes people whose output falls below standard. The trade-off is cost: skilled annotators earn far more than crowd workers, so the data is expensive. Labs pay because a smaller amount of excellent data can improve a model more than a large amount of mediocre data.

Revenue per employee

With roughly 110 staff and revenue above US$1 billion, Surge generated around US$10 million of revenue per employee in 2024. For comparison, large public software companies typically generate well under US$1 million per employee. The figure flatters Surge, because its contractors aren't counted as employees and much of the revenue passes through to them, but it still shows how lean the company's core team is.

Latest developments (October 2026)

Surge AI built a large data-labelling business without outside money, then explored its first raise. Reuters reported in July 2025 that it was seeking up to US$1 billion at a valuation above US$15 billion, according to TradingView (Reuters). Later reports suggested a higher figure, but we have not seen confirmation that a round closed, so our profile does not list a valuation. Surge supplies expert human data to leading AI labs. See our AI training data guide.

What to watch

Watch for a confirmed first funding round and its valuation, legal outcomes on contractor classification, revenue updates and changes in how AI labs source training data.

Takeaways

AI teams should judge data vendors on quality and security, not only price. Founders can see that bootstrapping to scale is possible even in AI. Investors should note that Surge's value rests on a small number of very large customers.

Sources

Every figure on this page comes from the public sources below. Where sources disagreed, we used the company's own announcement or a primary news report. Private-company data changes; check the date beside each figure.

1. Wikipedia