 | |
| Company type | Private |
|---|---|
| Industry | Information technology |
| Founded | 2016; 10 years ago (2016) |
| Founders | Alexandr Wang Lucy Guo |
| Headquarters | San Francisco, California,United States |
Key people | Jason Droege (CEO)[1] |
| Revenue | $870 million (2024)[2] |
| Owners | Meta Platforms (49%)[3] |
Number of employees | 1,200 (2025)[4] |
| Subsidiaries | Remotasks Outlier |
| Website | scale |
Scale AI, Inc. is an Americandata annotation company based inSan Francisco,California. It providesdata labeling,model evaluation, and software to develop applications for artificial intelligence.
The company’s research arm, theSafety, Evaluation and Alignment Lab, focuses on evaluating and aligninglarge language models (LLMs),[5] including through initiatives such asHumanity's Last Exam, a benchmark designed to assess advanced AI systems on alignment, reasoning, and safety.[6] Scale AI outsources data labeling through its subsidiaries,Remotasks, which focuses on computer vision and autonomous vehicles,[7] andOutlier, which focuses on data annotation for LLMs.[8]
Scale AI's customers in the commercial sector have includedGoogle,Microsoft,Meta,General Motors,OpenAI, andTime.[9][10][11][12] The company also directly works with world governments, including the United States on multiple military-related projects,[13] and with Qatar to improve the efficiency of its social programs.[14]
Scale was founded in 2016 byAlexandr Wang andLucy Guo throughY Combinator.[15] The pair previously worked together atQuora.[16] Initial investors of Scale includedDragoneer Investment Group,Tiger Global Management andIndex Ventures.[17][18][19] Lucy Guo was fired two years later in 2018.[20]
In August 2019, afterPeter Thiel’sFounders Fund made a $100 million investment in Scale, its valuation exceeded $1 billion and it acquiredunicorn status.[17][18]
Scale contracted with theUnited States Department of Defense in 2020.[21]
In May 2021,Michael Kratsios,Chief Technology Officer of the United States under theTrump administration, joined as Scale AI's managing director and head of strategy.[22]
By July 2021, Scale had reached a valuation of $7 billion, after a financing led by Greenoaks, Dragoneer Investment Group and Tiger Global Management.[23] There was an increased demand for data labelling from clients in different industries.[19]
In January 2022, Scale AI won a contract worth $250 million to give American federal agencies access to its suite of tools.[24] In February 2022, Scale AI developed its Automated Damage Identification Service in response to theRussian invasion of Ukraine. Satellite imagery was analyzed, measuring the damage to buildings, which were then geotagged and reported to humanitarian groups.[25] In November 2022, Scale AI was recognized byTime on it’s Best Inventions of 2022 list.[26] The company also opened an office in St. Louis in that same year.[17]
In January 2023, Scale laid off 20% of its workforce.[27]
In May 2023, Scale AI signed a deal with theUS Army’sXVIII Airborne Corps, becoming the first AI company to deploy its LLM (known as Donovan) on a classified network.[16]
In August 2023, Scale AI partnered with OpenAI, becoming the company’s "preferred partner" to fine-tuneGPT-3.5.[28] The company's services were used in the initial creation ofChatGPT.[17] In that same month, Scale AI’s evaluation platform was used atDEF CON, a hacking convention, at its first generative AIred team event, testing models provided by various companies.[29]
In December 2023, Scale AI was among a list of companies that contributed to Meta Platforms’s Purple Llama initiative, a security framework for the purpose of development of open generative AI models.[30]
In February 2024, Scale AI was selected by the Department of Defense to test and evaluate its LLMs for military purposes under a one-year contract.[31]
In March 2024, Scale reached a valuation of almost $13 billion after Accel led another round of funding.[32] In May 2024, Scale raised an additional $1 billion with new investors includingAmazon and Meta Platforms. Its valuation reached $14 billion.[33]
In August 2024, Scale signed an agreement with the USAI Safety Institute, collaborating with the agency on research, testing, and evaluation of the company’s AI models.[34] The US AI Safety Institute is controlled by theDepartment of Commerce’sNational Institute of Standards and Technology.[35]
In December 2024, Scale was sued by a former employee, alleging that the company was committing wage theft and misclassifying workers.[36] The following month, a second employee filed a similar suit.[37] In January 2025, several contractors sued Scale alleging psychological harm from being exposed to disturbing content.[38]
In January 2025, it was reported inThe Conversation that Scale AI and Meta had previously teamed up to create and sell Defense Llama, an LLM product with military-style defense purposes.[39] The company also took out a full-page ad inThe Washington Post, appealing to American PresidentDonald Trump to "win the AI war".[40] Later in the month, Scale AI and theCenter for AI Safety partnered to releaseHumanity's Last Exam, a benchmark test for AI systems.[6] The company has also assisted in the development of the benchmarks EnigmaEval, MultiChallenge, and MASK.[41]
In February 2025, Scale AI agreed to a five-year partnership with theQatari government to improve government services via AI-based tools and training, includingpredictive analytics,automation, and advanceddata analytics.[14] The deal was signed at the Web Qatar 2025 Summit byMohammed bin Ali bin Mohammed Al Mannai, theQatari Minister of Communications and Information Technology.[42] Also in February, the company became a third-party evaluator of AI models for the U.S. AI Safety Institute.[5]
In March 2025, Scale AI reached a deal with the United States Department of Defense to develop the Thunderforge project. The project aims to use AI to “plan and help execute movements of ships, planes, and other assets”, with the goal of speeding up military decisions in both peace and wartime. The contract was awarded to Scale AI and other companies (such asAnduril Industries andMicrosoft) by theDefense Innovation Unit, and is intended to first be used with theUSINDOPACOM andEUCOM.[13]
In April 2025, Scale AI released Scale Evaluation, a platform used to test LLMs against benchmarks to pinpoint weaknesses and flag where additional training data would improve the model.[41]
On June 10, 2025, it was reported thatMeta Platforms had agreed to purchase a 49% stake in Scale AI for $14.8 billion,[3] with the goal of accessing specialized datasets to improveLlama, a group of LLMs.[43] The company will remain as a standalone, independent entity from Meta.[1] Former CEO Alexandr Wang took a top position inside Meta as a part of the deal[3] and was replaced by the company's chief strategy officer and formerUber executive, Jason Droege.[44] Additionally,Google (which was Scale AI's largest customer) stated its intentions to cut ties with the company as a result of the deal.[45]
In 2017, Scale AI established Remotasks, acrowdworking platform to support the creation of labeled data for machine learning, particularly in areas such ascomputer vision andautonomous vehicles.[17][46][7] The subsidiary has facilities in Southeast Asia and Africa.[17][46]
In 2019, Scale AI set up a company called Smart Ecosystem Philippines to operate Remotasks within the country.[46] In the Philippines, many of Remotasks' hires are freelance contractors not covered under labor laws.[46] The pay for some annotation tasks dropped to less than one cent due to "vicious competition" after Remotask expanded to India as well as Venezuela.[46] Late payments are reportedly "commonplace", and some workers received only a few percent of their promised compensation.[46] In 2022, anOxford Internet Institute study said Remotasks met the "minimum standards of fair work" in only one out of ten criteria.[46]
Remotasks has been criticized for obscuring its affiliation with Scale AI, opaque communications, and abrupt changes in worker access in some regions.[17][46] In early 2024, the platform terminated operations in several countries, including Kenya, Nigeria, and Pakistan, citing administrative and operational considerations.[47]
Outlier is a separate contributor platform operated by Scale AI, designed for generative AI data work, particularly in the development and fine-tuning of LLMs.[8] Contributors on Outlier typically include professionals with advanced degrees, industry expertise, and native fluency in various languages.[48] Outlier tasks involve content evaluation andreinforcement learning from human feedback (RLHF).[49]
