Scale AI secures a $500 million contract with the Pentagon, five times larger than its previous agreement.

      The agreement was signed with the Chief Digital and Artificial Intelligence Office of the Pentagon, which is the main procurement body for the majority of US military AI integration investments planned for 2025 and 2026. Microsoft, Amazon, and Google also finalized similar classified-network agreements that same week.

      As reported by Bloomberg today, Scale AI has secured a $500 million contract from the US Department of Defense, marking one of the largest individual AI procurement contracts awarded by the Pentagon and a fivefold increase in the company's existing collaboration with the department. This contract is intended to facilitate the incorporation of AI tools into military decision-making and data-processing workflows.

      This new contract is five times larger than the $100 million deal Scale was given in September 2025. Dan Tadross, head of Scale’s public-sector business, mentioned to Bloomberg that the Pentagon was already "pushing the limits" of the initial agreement. This expansion aligns with the trend of increasing AI procurement spending by the Department of Defense, expected to escalate through 2025 and 2026.

      The contract was executed with the Pentagon’s Chief Digital and Artificial Intelligence Office (CDAO), which is tasked with enhancing AI adoption in the military’s operational and support functions. CDAO has emerged as the primary procurement entity for large-scale AI integration in 2025 and 2026. Scale AI is joining a select group of vendors that are seeing significantly increased commitments from CDAO.

      In May 2026, Microsoft, Amazon, and Google also entered into significant classified military-use AI agreements. The Pentagon has approved the AI systems of eight firms for use on classified networks, including major cloud providers. The $500 million contract awarded to Scale AI fits within this larger growth but is strategically distinct; while the contracts with hyperscalers mainly focus on cloud and model infrastructure, Scale’s responsibility is in the data-labelling and decision-support layer that functions above that.

      Scale AI's defense sector is its fastest-growing segment. The company has concentrated on addressing the data-quality bottleneck that has impeded the effective deployment of AI within military applications. While AI models excel in benchmarks, their performance diminishes when the underlying training data is fragmented, mislabelled, or inconsistent with real-world operations. Scale's value proposition lies in enhancing that data layer, which is essential for dependable deployment.

      Scale's designation, backed by Meta, reflects the social media company's investment in June 2024, which, along with other strategic funding, valued Scale at over $14 billion. This connection with Meta provides Scale with the financial resources necessary to expand its public-sector operations at the pace demanded by the new contract. Scale’s CEO, Alexandr Wang, has also become a prominent advocate for US AI policy, frequently testifying before Congress on the nation's military AI competitiveness.

      As Scale's public-sector growth has progressed, its broader strategic position has also evolved. The company is venturing into what is increasingly being referred to as “physical AI” applications, where AI systems are implemented in autonomous platforms, robotics, and uncrewed military hardware. The data-labelling foundation offers the same value proposition for these platforms as it has for conventional military AI processes. The new $500 million contract provides Scale with the financial latitude to expand in both areas simultaneously.

      Three implications can be drawn from this announcement. Firstly, the Pentagon's AI procurement budget has now grown large enough to accommodate multiple competing vendors at the half-billion-dollar single-contract level, moving away from the smaller awards that were prevalent as recently as 2024.

      Secondly, the data-labelling and decision-support layer that Scale occupies has been recognized as a distinct procurement category, separate from cloud and model infrastructure, complete with its own budget and competitive dynamics.

      Lastly, the Meta investment in Scale, initially described as a partnership to enhance AI training-data quality, has resulted in a public-sector commercial outcome significantly larger than the initial commercial application.

      However, the contract does not yet address the longer-term question of whether the data-labelling layer will remain a separate category as foundational models continue to advance, or if large model providers will integrate that function into their offerings. The $500 million contract provides Scale with approximately two to three years of operating room to resolve that issue on its own terms. Currently, the broader Pentagon AI procurement cycle appears to be progressing rapidly enough that...