US Navy bans use of DeepSeek AI over security and ethical concerns

The US Navy has prohibited its personnel from using DeepSeek AI, citing security and ethical risks tied to the model developed by the Chinese startup.

In an internal email sent Friday, the Navy directed members to avoid using DeepSeek “in any capacity”, warning of potential threats related to the model’s origin and usage. 

A Navy spokesperson confirmed the memo’s authenticity, linking it to the Department of the Navy’s Chief Information Officer’s generative AI policy.

The directive follows the release of DeepSeek R1, a powerful open-source AI model that has surged past OpenAI’s ChatGPT in App Store downloads. Industry analysts have praised its performance and reasoning capabilities, sparking speculation that AI infrastructure costs could drop significantly.

DeepSeek claims it built R1 in just two months for under US$6 million, despite US restrictions on advanced chip exports to China. By comparison, US AI giants such as OpenAI, Google, and Anthropic have spent billions developing similar models.

The shockwave from DeepSeek’s rapid rise hit markets hard, with AI chipmakers Nvidia and Broadcom each plunging 17% on Monday, erasing a combined US$800 billion in market value. The Nasdaq dropped 3.1%, reflecting broader fears about shifts in AI investment.

The Navy’s warning was issued before the market turmoil. The email, sent to the Operational Navy (OpNav) distribution list, advised personnel that DeepSeek should not be used for work-related tasks or personal activities. The advisory originated from the Naval Air Warfare Center Division Cyber Workforce Manager, urging recipients to avoid downloading or installing the model.

Meanwhile, DeepSeek temporarily halted new user registrations on Monday, citing large-scale cyberattacks on its platform before resuming operations.

President Donald Trump, who took office last week, called DeepSeek’s rise “a wake-up call” for US tech companies. His administration has taken a mixed stance on Chinese technology, with Trump currently working to keep TikTok operational in the US, despite previous efforts to ban it.

David Sacks, Trump’s AI and crypto czar, emphasized the high stakes of the AI race. 

“DeepSeek R1 shows that AI will be highly competitive,” he posted on X, adding, “We can’t be complacent.”

In response, Meta reportedly launched four DeepSeek-related “war rooms” within its AI division, reflecting growing concerns about China’s progress in the field.

Industry leaders are also weighing in. Scale AI CEO Alexandr Wang called DeepSeek’s advancements “earth-shattering”, saying R1 is “on par with the best American models”. He described the US-China AI race as an ‘AI war’, highlighting the need for continued US investment.

Last week, the Trump administration announced Stargate, a joint AI infrastructure initiative between OpenAI, Oracle, and SoftBank, aimed at strengthening America’s AI dominance.

Share this Post:

Accessibility Toolbar