DeepSeek, a Chinese AI company, has successfully raised $15 billion in a recent funding round as it aims to compete with major US labs like OpenAI and Anthropic, according to reports.
Based in Hangzhou, this firm is among several open-source rivals enticing users with affordable and high-performance AI models. Interestingly, they had to increase their fundraising target not once but twice due to overwhelming interest from investors, as reported by Bloomberg, citing sources familiar with the situation.
Among the investors in this funding round are tech giant Tencent and battery manufacturer Contemporary Amperex Technology, as per the same report.
There’s talk that DeepSeek is gearing up for an initial public offering in Shanghai in the near future.
Back in January 2025, DeepSeek’s launch of an open-source model sent shockwaves through the market, contributing to a staggering $1 trillion drop, as it offered features comparable to those of leading US labs.
Other Chinese startups, like Moonshot and Z.AI, have also launched their own open-source models recently, leading to concerns that China might be closing the gap with the US in the realm of advanced AI development.
As previously mentioned, there are significant concerns about DeepSeek’s data management practices, which involve storing information on servers located in China. This situation mirrors the national security issues that led to Congress banning TikTok.
Earlier this week, federal agencies were instructed by the Department of Homeland Security to eliminate DeepSeek’s models from their systems.
DHS highlighted that these models pose various security threats and cautioned that their use could result in transferring sensitive data back to the People’s Republic of China, especially if utilized by US government personnel or contractors.
In addition, the US Navy, as well as the Commerce Department and New York state, have initiated similar restrictions regarding DeepSeek’s models.
Meanwhile, both OpenAI and Anthropic have publicly accused DeepSeek of employing a method known as “distillation” — using advanced AI models for training lesser ones, without permission, which significantly lowers the cost and time involved in product development.
In April, the White House’s Office of Science and Technology published a memo charging that “foreign entities, principally based in China,” are engaged in widespread theft campaigns of intellectual property.
