Scale AI

Scale AI, founded in 2016 by Alexandr Wang and Lucy Guo, is a San Francisco-based company that specialises in data annotation, model evaluation, and software solutions for artificial intelligence (AI) applications. The company's primary mission is to provide high-quality data and technologies that enable businesses and governments to deploy AI systems effectively. 

Scale AI serves various industries, including automotive, government, and generative AI, by offering advanced data and AI solutions. It is known for its Scale Data Engine, which aids in training models, and the Scale GenAI Platform, which provides full-stack generative AI capabilities.
 

The company also runs SEAL (Safety, Evaluation, and Alignment Lab), which conducts rigorous testing and evaluations of AI models, ensuring they are reliable and safe. Through Scale Donovan, the company provides mission-critical agentic AI solutions for diverse sectors, including the military, where they work with organisations like the US Army and Air Force.
 

Scale AI has partnerships with industry leaders such as Meta, Google, Microsoft, and OpenAI, and collaborates with governments, including the US and Qatar, to implement AI solutions across various public sector programs. In 2024, Scale AI reached a valuation of $14 billion after raising additional investments from companies like Meta and Amazon. The company also developed solutions in response to global challenges, such as using AI for damage assessment during the Russian invasion of Ukraine.
 

Throughout its growth, Scale has expanded its operations globally and made key acquisitions, including Remotasks and Outlier.ai, which support the company’s data labelling needs. Scale AI continues to drive AI innovation, working at the frontier of model safety and generative AI, positioning itself as a leader in the AI technology space.

Scale AI was founded in 2016 by Alexandr Wang and Lucy Guo, both former employees of Quora. They started the company with the goal of providing high-quality data and software solutions to help businesses and governments implement artificial intelligence (AI) systems. 

The company initially gained attention through its partnership with Y Combinator, which helped launch and fund the company. Scale AI’s primary service was data annotation, which involved labelling data for machine learning models. This allowed AI systems to learn from labelled data and improve their accuracy.
 

In its early years, Scale AI grew rapidly, attracting major investors such as Dragoneer Investment Group, Tiger Global Management, and Index Ventures. The company quickly expanded its offerings beyond data labelling to include AI model evaluation and software tools to support the development and deployment of AI applications. 

Scale AI also focused on creating a research arm, called the Safety, Evaluation, and Alignment Lab (SEAL), to improve AI safety and model alignment with human values. In 2019, Scale AI reached a significant milestone when it raised $100 million in funding from Peter Thiel’s Founders Fund, which increased the company's valuation to over $1 billion, earning it “unicorn” status.
 

As the demand for AI services increased, Scale AI further expanded its client base to include top companies like Google, Microsoft, Meta, and General Motors, as well as government entities like the United States Department of Defence. In 2021, Scale AI’s valuation reached $7 billion after another round of funding. The company’s services were also used to support humanitarian efforts during the Russian invasion of Ukraine in 2022, where it analysed satellite imagery to assess the damage to buildings.

Scale AI’s focus on large-scale AI applications continued to grow. In 2023, the company became a key partner of OpenAI to fine-tune GPT-3.5 and further solidified its position in the AI industry. It also signed a deal with the U.S. Army to deploy its AI models on classified networks. By 2024, Scale AI’s valuation reached $14 billion, and the company was actively working with the U.S. government on several military-related projects.

In 2025, Scale AI partnered with Meta Platforms, with Meta acquiring a 49% stake in the company for $14.8 billion. This partnership aimed to improve Meta’s generative AI models, specifically Llama, and it led to leadership changes within the company. 

Scale AI continues to work with various government agencies, including the U.S. Department of Defence, and supports commercial enterprises by offering advanced data solutions for AI systems. Scale AI’s products and services, such as the Scale Data Engine and Scale GenAI Platform, continue to be used across industries, including defence, automotive, and healthcare.
 

The company’s recent projects include the development of Thunderforge, an AI-powered project aimed at improving military decision-making and the deployment of AI solutions for public sector services in Qatar. Scale AI’s ongoing work focuses on further improving AI model evaluation, safety, and alignment, with the company positioning itself as a leader in AI technology. Today, Scale AI has grown to employ over 1,200 people and continues to shape the future of AI through its advanced data solutions and partnerships.

Scale AI’s mission is to develop reliable and high-quality AI systems that can support the world’s most important decisions. The company aims to provide advanced data solutions and technologies that power the most sophisticated AI models used by businesses, governments, and organisations. By focusing on data annotation, model evaluation, and AI safety, Scale AI ensures its systems are accurate, effective, and aligned with human values. Its goal is to make AI more accessible and trustworthy, helping clients across various sectors such as healthcare, defence, and finance use AI to solve real-world challenges and create lasting impact.

Scale AI’s vision is to develop reliable AI systems that can make a real impact in the world’s most important decisions. The company aims to provide high-quality data and cutting-edge technologies to help businesses, governments, and organisations build and deploy AI applications. By focusing on creating advanced AI agents and ensuring their safety and alignment with human values, Scale AI seeks to lead the development of AI systems that are both innovative and trustworthy, ultimately making AI more accessible and effective for various industries across the globe.

Scale AI has been recognised for its significant contributions to the AI field. In 2022, the company was featured in Time’s Best Inventions of 2022 for its work in AI and data solutions. The company has also been acknowledged for its impactful partnerships with major organisations like Meta, Google, and the U.S. Department of Defense. Scale AI has earned a reputation for its excellence in data annotation, model evaluation, and safety.

Scale AI provides a range of products and services that help businesses, governments, and organisations develop, deploy, and manage artificial intelligence (AI) systems. The company focuses on high-quality data solutions, model evaluation, and AI safety, which are essential for building reliable and effective AI applications. Below are the key products and services offered by Scale AI:
 

Scale Data Engine- The Scale Data Engine is a comprehensive platform designed to help organisations generate and manage high-quality datasets for training AI models. This platform enables businesses to integrate their own data into machine learning systems, providing the base for AI solutions that can drive strategic differentiation. It supports a wide variety of data types, including text, images, video, and more, making it suitable for various industries, from healthcare to automotive. The data engine ensures the data is clean, accurate, and ready to train AI models, enhancing their performance and reliability.

Scale Generative AI Platform- Scale's Generative AI Platform allows customers to build, evaluate, and control advanced AI agents and applications that continuously improve. This platform is especially valuable for organisations looking to develop AI models that can adapt and learn over time through human-agent interactions. It supports large-scale AI projects, from fine-tuning foundation models to deploying generative models for specific business needs. The platform is designed to be flexible, allowing businesses to create custom solutions tailored to their unique data and objectives.
 

Scale Donovan- Scale Donovan is an AI solution designed to assist in complex decision-making tasks across various sectors, particularly in mission-critical areas such as defence and public safety. It is used by organisations like the U.S. Army to support operations with AI-powered agents that can reason, plan, and take action based on available data. Donovan is a part of Scale AI's broader efforts to integrate AI agents into real-world applications where rapid decision-making is crucial. It helps reduce human intervention while improving efficiency and effectiveness in complex environments.
 

Safety, Evaluation, and Alignment Lab (SEAL)- The SEAL research arm of Scale AI is dedicated to improving AI model safety, evaluation, and alignment with human values. SEAL conducts rigorous testing of AI models to ensure they are not only effective but also trustworthy. One of the most notable benchmarks created by SEAL is Humanity’s Last Exam, which assesses AI systems on alignment, reasoning, and safety. This lab helps organisations understand the strengths and weaknesses of their AI models, ensuring that they meet high standards for performance and reliability. It is a key part of Scale AI's efforts to make AI systems safer for real-world applications.
 

Data Annotation Services- Scale AI offers data annotation services through its subsidiary Remotasks. This service supports the creation of labelled data used to train machine learning models, particularly for computer vision and autonomous vehicles. Remotasks employs a global workforce to label large volumes of data accurately, helping organisations quickly prepare datasets for model training. These annotated datasets are crucial for improving the accuracy of AI systems in tasks like image recognition, object detection, and facial recognition.
 

Outlier AI- Outlier is another service offered by Scale AI that focuses on data annotation for large language models (LLMs). It provides data labelling services for LLMs, helping businesses fine-tune their models by providing high-quality content evaluation. Outlier specialises in reinforcement learning from human feedback (RLHF), which improves AI models' ability to understand and process complex data inputs. This service is especially valuable for companies building conversational AI systems, chatbots, or other language-based applications.
 

AI Safety Solutions for Government and Military- Scale AI works with government agencies and military organisations to enhance AI-driven solutions for national security and public safety. The company provides customised solutions for various government departments, including the Department of Defense and Intelligence Community. Scale AI’s services help these agencies improve operations, decision-making, and security by using advanced computer vision algorithms and agentic AI systems. For example, Scale AI is working on projects like Thunderforge, which aims to use AI for military planning and decision-making.