The Department of Defense has awarded Scale AI a contract to integrate AI agents into military decision-making processes, signaling a major step forward in the application of artificial intelligence to defense. The program, known as Thunderforge, aims to enhance the speed and effectiveness of military operational planning. This development underscores the growing ties between Silicon Valley and the military.
Scale AI, a data training startup valued at $13.8 billion, will provide its expertise to aid the DoD. The company supplies data to major tech firms such as OpenAI, Google, Microsoft, and Meta. The contract marks a significant shift toward AI-powered, data-driven warfare, enhancing the ability of U.S. forces to anticipate and respond to threats.
Chris Morales, a partner at Point72 Ventures, noted in an email that the DoD’s embrace of AI suggests a recognition that AI-driven decision-making tools can significantly influence the outcomes of conflicts. The Defense Innovation Unit (DIU) awarded the prototype contract for the Thunderforge program, which plans to integrate AI into military operations. The project will involve partnerships with companies like Anduril and Microsoft, to develop and deploy AI agents.
According to the DIU, Thunderforge’s technology will offer AI-assisted planning, decision support tools, and automated workflows, ensuring that U.S. forces can respond to threats with speed and precision. The partnership is further evidence of the deepening relationship between Silicon Valley and the military, driven by the vast potential of AI within the defense sector.
In 2024, global venture capital investments in defense-related companies surged, reaching $31 billion, according to McKinsey. The deal follows recent announcements from defense tech companies. Shield AI, which focuses on autonomous aircraft and AI-powered software, recently announced a $240 million funding round, valued at over $5 billion.
Chip Walter, a partner at Marlinspike, a firm focused on defense tech added that this represents “a great sign for the DoD to be embracing a software approach to the ever-growing interest in and utilization of AI.” He emphasized that a program of the scale and importance of Thunderforge helps solidify the defense relationship for companies like ScaleAI.
Morales also pointed out that defense-focused AI startups and established IT firms, who possess skilled AI engineers, are well-positioned to collaborate with the DoD on AI-driven decision-making. There is optimism among investors regarding the Pentagon’s rapid adoption of AI technologies.
However, AI and military researchers are also focused on ensuring that AI is integrated into military operations safely and effectively. Yasir Atalan, a data fellow at the Center for Strategic and International Studies (CSIS), noted that many military workflows are ripe for AI integration. These workflows include action generation, logistics, combat readiness assessments, and war gaming, where AI can potentially improve efficiency and effectiveness.
At CSIS, Atalan’s research includes examining potential biases in off-the-shelf AI models. He studies whether these AI models demonstrate particular tendencies, which could include escalatory, aggressive, or cooperative biases, when evaluating AI adoption in the military. “These are the questions we need to always, constantly, be working on,” he said.
The Scale AI-DoD partnership also raises questions about human oversight in AI-driven military operations. While Scale AI’s statement emphasized that their technology would always operate under human oversight, the DIU’s release did not. Atalan emphasizes the key goal is to responsibly integrate AI into military planning. He acknowledges that because AI will increase the speed of decision making, the military’s ability to adapt is critical.
