US spy agencies race for Nvidia AI chips under new $9 billion plan
The White House approved a $9 billion plan to accelerate U.S. intelligence agencies' adoption of advanced AI chips, including Nvidia’s Grace Blackwell superchip, amid fears of falling behind rivals and private AI labs. The initiative also involves $800 million for faster computing capacity acquisition, while tensions persist over military use of AI models like Anthropic’s Mythos, despite ongoing contracts with the NSA.
The White House has approved a $9 billion funding initiative to help U.S. intelligence agencies rapidly integrate advanced AI capabilities, focusing on acquiring cutting-edge computer chips like Nvidia’s Grace Blackwell superchip. The move comes as agencies such as the National Security Agency (NSA) and the Department of Defense (DoD) face a shortage of AI chips critical for next-generation surveillance, cyberwarfare, and intelligence analysis. The Trump administration is also considering an additional $800 million to speed up computing capacity acquisition, reflecting concerns that the U.S. is falling behind both private AI labs and geopolitical rivals. The funding aligns with broader efforts to treat AI as a national security asset, comparable to nuclear technology or satellite systems. U.S. government agencies currently run AI models on Amazon Web Services, which recently announced a $50 billion upgrade for government cloud computing services. Meanwhile, top AI firms—OpenAI, Anthropic, and xAI—have agreed to provide pre-release access to advanced models for government vetting, though tensions remain. Anthropic initially resisted military and surveillance use of its technology, leading to a clash with the Pentagon over demands for unrestricted ‘lawful use’ of its systems. Despite these disputes, the NSA has finalized a contract to continue using Anthropic’s Mythos model, despite the Pentagon designating the company a supply chain threat. The decision follows authorization from Susie Wiles, the White House chief of staff, to bypass restrictions. Earlier in 2026, Trump abruptly canceled a planned AI executive order hours before signing, citing objections from figures like Elon Musk, Mark Zuckerberg, and former AI adviser David Sacks. Trump had previously revoked a major AI safety executive order issued by Joe Biden shortly after taking office in January 2025, arguing it could hinder U.S. leadership over China. The initiative underscores the U.S. government’s urgency in securing AI dominance, with Vinh Nguyen, a former NSA chief data scientist, emphasizing the need for ‘frontier AI chips, models, systems, and talent’ to match evolving threats. The $9 billion request now awaits congressional approval, marking a pivotal step in militarizing AI infrastructure.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.