US Adds Four AI Firms to Defense Roster, Eyes Anthropic
The US Department of Defense has expanded its list of approved AI providers by adding four companies. Agreements with Microsoft, Reflection AI, Amazon, and Nvidia now permit their technologies in classified military operations. These firms join OpenAI, xAI, and Google, enabling deployment for any lawful purpose.
Reflection AI has not yet released a public model. OpenAI, known for ChatGPT since its 2022 launch by former OpenAI researchers including Sam Altman, leads in generative AI. xAI, founded by Elon Musk in 2023 as a rival to OpenAI, focuses on understanding the universe through AI. Google, with its DeepMind division acquired in 2014, develops advanced models like Gemini.
Dispute Centers on Anthropic's Stance
A recent conflict highlighted differences with Anthropic. CEO Darius Amodei objected to the "any lawful use" clause. He argued it could enable surveillance of US civilians and development of autonomous weapons, uses Anthropic sought to avoid. The Pentagon ended a $200 million contract. Anthropic filed a lawsuit, seeking compensation for lost government revenue and deals affected by the decision.
The Trump administration labeled Anthropic a "supply chain risk," a first for a domestic firm. Officials called the company "woke." Anthropic, founded in 2021 by ex-OpenAI executives including Dario Amodei, emphasizes AI safety with its Claude models.
Pentagon Builds Flexible AI Architecture
The Pentagon's announcement stresses avoiding dependence on single vendors. It aims for long-term options for joint forces. The tools will equip soldiers to respond confidently and protect against threats. Applications cover Impact Level 6 for secret data and Level 7 for top-secret materials, fostering an "AI-first fighting force."
Stay ahead of the AI curve
The most important updates, news, and content — delivered weekly.
No spam. Unsubscribe anytime.
Current generative AI serves unclassified tasks like drafting documents, summarizing, and research across defense agencies. New providers will assist in combining data, improving awareness, and supporting decisions in tough settings. Details on domestic use remain unclear.
Past employee protests at Google and Amazon led to firings over military AI involvement. Widening the supplier base reduces risks from company leaders' shifts.
Anthropic's Claude powered classified work in Palantir's Maven system, a role newcomers might assume. Its Mythos model aids the National Security Agency in cyber defense and warfare. Globally, 40 organizations evaluate Mythos; 12 are public, with UK MI5 and US NSA likely among the rest.
Signs of Reconciliation Emerge
Axios reports the administration may soften its position. A White House source said efforts aim to "save face and bring 'em back in." Anthropic's Claude coding model continues in US security agencies despite the dispute.
The White House states it actively collaborates with government and industry, including frontier AI developers, to safeguard the nation and citizens.
Microsoft, a key OpenAI backer with Azure cloud, Amazon via AWS services, and Nvidia with dominant GPUs for AI training, strengthen defense capabilities. This move diversifies sources amid growing military AI needs.