This week in Washington, US lawmakers are pushing for a national ban on DeepSeek, the Chinese generative AI tool that stirred controversy and caused significant losses for US tech companies back in January.
The No DeepSeek on Government Devices Act, introduced by Democratic Congressman Josh Gottheimer from New Jersey and Republican Congressman Darin LaHood from Illinois, aims to prohibit DeepSeek’s software on government devices. Countries like Australia and Taiwan have already blocked the tool due to national security concerns. Italy has imposed restrictions on how DeepSeek handles its users’ data, and more nations are weighing similar actions.
In the US, Texas has moved to limit DeepSeek usage on state-owned devices, and even the US Navy has imposed restrictions. The situation is reminiscent of the ongoing debate around TikTok. The government fears that the Chinese Communist Party could exploit user data through this platform, echoing longstanding worries about Chinese tech companies.
LaHood voiced his concerns clearly: “DeepSeek, linked to the CCP, poses a national security risk. This legislation aims to keep sensitive data out of their reach and protect Americans.” Gottheimer added, “The CCP is determined to undermine our security. We have evidence they’re using DeepSeek to access US citizens’ sensitive data. We can’t let them infiltrate the devices of government officials.”
On a parallel track, Republican Senator Josh Hawley from Missouri introduced the Decoupling America’s Artificial Intelligence Capabilities from China Act. This legislation targets the importation of tech developed in China, imposing severe penalties for violations, including up to 20 years in prison and hefty fines. Though it doesn’t specifically mention DeepSeek, it reflects growing unease about foreign technology in government and military areas.
Mel Morris, CEO of AI research engine Corpora.ai, highlighted a long-standing anxiety about non-allied technology, citing Huawei as an example. He believes the ban on DeepSeek may be a wise move to prevent possible compromises during conflicts.
Ilia Kolochenko, CEO of ImmuniWeb, urged lawmakers to broaden their focus. He pointed out that while DeepSeek’s risks are significant, many other generative AI vendors may pose similar or even greater threats. Some of these companies have used copyrighted materials to train their models without permission, creating a different set of ethical and legal concerns.
Kolochenko warned against allowing the DeepSeek incident to overshadow the broader issues faced by other AI companies.