Scale AI Secures Record Investment from Meta, Validating Lotus's Foresighted Strategy
On June 10, Meta announced a record $14.8 billion investment to acquire 49% equity in Scale AI – marking one of the largest private company financings in history. This strategic investment, comprising 20% of Meta’s annual AI budget, not only underscores Scale AI’s irreplaceability as an industry-leading data infrastructure provider but also validates Lotus’s long-held insight that "data is the vital lifeline of AI."
As part of the transaction, Scale AI founder Alexandr Wang will assume a senior role at Meta leading a new "Superintelligence Lab," driving development of next-gen AI models and data ecosystems. This is rare for Mark Zuckerberg, who traditionally promotes internal loyalists to leadership positions. This move demonstrates both Scale AI's competitiveness and Zuckerberg’s conviction that external visionaries like Wang can best fortify Meta’s AI leadership.
With AI being Meta’s top 2025 priority, the lukewarm reception of April’s Llama 4 model amid intensifying industry competition has heightened urgency. Zuckerberg deeply admires Wang’s unique ability to navigate AI’s technical complexities while building commercially viable ventures beyond pure research – key to realizing Meta’s AI ambitions.
Founded in 2016, Scale AI specializes in preparing training data for cutting-edge LLMs, serving tech giants including OpenAI, Google, and Meta – its largest customer. Recent collaborations with Meta extend into defense technology, where both companies co-develop "Defense Llama," a customized version of Meta’s open-source model tailored for U.S. national security missions.
"Scale has annotated roughly 70% of all major AI models," noted Vahan Petrosyan, CEO of rival SuperAnnotate. "By bringing Wang and his team, Meta essentially acquires the collective intelligence for building superior ChatGPT competitors."
Meta CTO Andrew Bosworth emphasized: "Scale’s technological breakthroughs and data assets will accelerate our Llama ecosystem. Wang’s mastery of data value injects new momentum into Meta’s AI strategy – this transcends financial synergy into deep technical integration."
Scale AI's Strategic Expansion
In its early days, Scale focused more on labeling images of vehicles, traffic lights, and road signs to assist in training models for autonomous driving. Since then, through reinforcement learning intelligent scheduling algorithms, dynamic workflow orchestration toolkits, and an "annotation-validation-iteration" closed-loop system, it has increased data processing efficiency 17 times that of traditional models, establishing full-stack service capabilities covering text, images, speech, and multimodal battlefield intelligence. It has gradually expanded into an enterprise-level data integrator, assisting clients in AI model development and deployment.
Scale AI has been strengthening strategic cooperation with governments worldwide:
In March, Scale AI announced it had secured the U.S. Department of Defense's flagship "Thunderforge" project contract, worth millions of dollars.
In April, Scale AI announced it had obtained $30 million in funding from the Canadian government to promote high-impact AI projects in the country, focusing particularly on supply chains, productivity enhancement, and AI commercialization.
In April, Scale AI announced a five-year cooperation agreement with the Qatari government, aiming to use AI technology to enhance multiple sectors in Qatar, including education, healthcare, transportation, tourism, and public services. Specific projects include developing customized AI learning platforms, AI teaching assistants, contract drafting and medical optimization, and traffic accident monitoring. Scale AI also stated that new collaborations with Asian and European governments could account for a significant portion of its sales in the next two quarters.
Alexandr Wang has proposed suggestions such as establishing a "National AI Data Repository" and ensuring power supply for data centers at congressional hearings, which have won the approval of bipartisan lawmakers. Michael Kratsios, a former senior executive of the company, is now a core tech advisor to Donald Trump.
Despite facing scarcity in traditional text data, Scale AI continues to evolve. It has shifted its labeling focus to high-expertise fields such as healthcare and law—for example, enhancing AI capabilities to handle tax law differences across countries.
Scale AI expects revenue to exceed $2 billion in 2025, doubling from $870 million in 2024, demonstrating strong growth momentum. As of the end of last year, the company had more than $900 million in cash on its balance sheet. In May 2024, Scale AI completed its Series F financing, led by Accel with follow-on investments from Meta, Amazon, etc., valuing the company at approximately $14 billion post-financing.
Data is the key to AGI. The cooperation between Scale AI and Meta will redefine the value of data in the AI era. We look forward to Scale AI continuing to drive technological innovation in data labeling while helping Meta build more robust AI infrastructure to create smarter experiences for global users.
Source： Scale AI, Reuters, CNBC