Pentagon: OpenAI and Google Lead in Military AI

AI, a Strategic Issue for the Pentagon

Artificial intelligence is becoming an essential technology for the U.S. armed forces. The Pentagon has decided to collaborate with nearly all the major American players in the sector to integrate these cutting-edge technologies into its operations. This initiative is set against a backdrop of global strategic competition, where the United States and China are striving to incorporate AI into their military ecosystems. A significant decoupling in the field of artificial intelligence is beginning to take shape.

Major Contracts with AI Giants

To maintain their supremacy, the United States has signed contracts with several major companies. Among them are NVIDIA, Amazon, Microsoft, Google, OpenAI, SpaceX (the parent company of xAI), Oracle, and Reflection AI. These collaborations aim to equip the armed forces with a "suite of artificial intelligence tools," according to Emil Michael, Under Secretary of Defense for Research and Engineering. These tools are designed to provide a decisive advantage on the battlefield and enable soldiers to achieve absolute decision-making superiority.

Specific Contributions from Companies

Each of the involved companies brings unique expertise. The AI models developed by OpenAI and Google will be utilized in classified operations. Microsoft, Amazon, and Oracle will provide their infrastructures, thus sparing the Pentagon from having to build its own. NVIDIA and Reflection AI, supported by NVIDIA, will offer open-source models that are more easily modifiable for creating defense tools tailored to the Pentagon's requirements.

Notable Absence of Anthropic

Among the major names in the sector, Anthropic stands out for its absence. This startup, despite its rapid growth, was not selected due to disagreements with the Pentagon over certain uses of AI, particularly mass surveillance and autonomous weapons. Anthropic had previously participated in operations in Venezuela and against Iran. Other signatories of the agreements have also expressed reservations about these controversial applications, refusing to allow their models to be used for such purposes. This situation highlights the ideological tensions surrounding the use of AI in the military domain.