US bans DeepSeek AI products at civilian federal agencies

    The U.S. Department of Homeland Security has banned the use of DeepSeek products at civilian federal agencies and in their information systems. According to a directive published in the Federal Register, products already installed on these systems must also be removed. The decision came several days after the Chinese company released its new artificial intelligence model. The restriction applies to all civilian federal agencies.    Under the new rule, agencies such as the Environmental Protection Agency and the Federal Communications Commission must stop using DeepSeek products. Related software on their systems must also be removed. DeepSeek will no longer be able to participate in federal government contracts. The ban applies to both software and services used by government agencies.    The department cited the possibility that sensitive data belonging to U.S. government employees and contractors could be transferred to the Chinese government as the main risk. The directive also states that DeepSeek models may contain malware or other security vulnerabilities. The department therefore considers keeping these products on government information systems dangerous.    Several days before the decision, DeepSeek had released its new V4.1-Flash-1017 model. According to the company, the model is smarter, faster and more efficient. The growing competition that its cheaper models pose to ChatGPT and Gemini has brought the company to the attention of the U.S. government. Last year, 2 members of the U.S. Congress introduced a bill to ban the DeepSeek chatbot on government devices.