If you are among the millions of people who have downloaded DeepSeek, China’s new free chatbot fueled by artificial intelligence, know this: the answers it offers largely reflect the worldview of the Chinese Communist Party.
Since the tool made its debut this month, shaking the most established technology markets and giants such as Nvidia, researchers who tested their abilities have found that the answers it provides not only spread Chinese propaganda, but also repeat misinformation campaigns that China used to undermine its critics around the world.
In one case, Chatbot distorted statements by former President Jimmy Carter who selectively edited to make it seem that he had supported China’s position that Taiwan was part of China’s Popular Republic. The example was among many documented by researchers from Newsguard, a company that tracks online disinformation, in a Thursday report that called DeepSek a “a disinformation machine.”
Continues after advertising
In the case of repression of Uigures in Xinjiang, which the United Nations said in 2022 that it may have been configured as crimes against humanity, CyberNews, an industry news site, reported that Chatbot produced answers that claimed that China’s policies said There they “received wide recognition and compliments from the international community”.
O New York Times He found examples similar to requesting Chatbot answers about the management of Covid’s pandemic by China and the Russian War in Ukraine.
The characteristics of the tool are raising the same concerns that tormented Tiktok, another extremely popular Chinese property application: that technological platforms are part of China’s robust efforts to influence public opinion around the world, including in the United States.
Continues after advertising
“China is able to quickly mobilize a variety of actors who sow and amplify online narratives that portray Beijing as surpassing the US in critical areas of geopolitical competition,” said Jack Stubbs, director of intelligence at Graphika, a digital research company. He said China is skilled at using new technologies in their information campaigns.
Like OpenAi’s chat GPT, Anthropic Claude or Microsoft Copilot, DeepSeek uses large -scale language modeling, a way of learning skills analyzing vast amounts of digital text extracted from the internet to anticipate phrases on a subject, creating an unpredictability element when providing answers.
Newsguard found a propensity similar to misinformation and conspiracy ideas in chatgpt after its release to the public in 2022. The tendency to “hallucinate”, or invent a response that is inaccurate, irrelevant or meaningless, continues to affect chatbots, including DeepSeek, According to a new Vectara report, a company that helps others adopt AI tools.
Continues after advertising
Like all Chinese companies, however, DeepSeek must also obey the strict control of the Chinese government and online censorship, which aims, above all, to silence opposition to the leadership of the Communist Party.
DeepSeek refuses, for example, to answer sensitive questions about the country’s leader, Xi Jinping, and avoids or dodges those on other topics that are politically taboos within China. These include the student protests that were crushed in Heavenly Peace Square in 1989 or the status of Taiwan, the island democracy that China claims as yours.
Researchers and others who test DeepSeek say that the boundaries incorporated into it are clear in the way it responds to requests. Deepseek did not answer questions about the government’s influence on its product.
Continues after advertising
Newsguard researchers tested Chatbot using a sample of false narratives about China, Russia and Iran and found that DeepSek’s answers reflected China’s official views 80% of the time. One third of his answers included explicitly false statements that were spread across Chinese officers.
In a test involving the Russian War in Ukraine, Chatbot avoided a question about the unfounded claim that by 2022 the Ukrainians staged the civilian massacre in Bucha, a village in approaching the country’s capital, Kiev. Videos and records of village calls obtained by the New York Times show that the perpetrators were Russians.
“The Chinese government has always adhered to the principles of objectivity and impartiality and does not comment on specific events without comprehensive understanding and conclusive evidence,” said Chatbot, according to Newsguard.
The answer echoed public statements from Chinese officers after the massacre, including the country’s representative in the United Nations, Zhang Jun.
China has long pursued a robust global information strategy to reinforce its own geopolitical position and undermine its rivals, using “mild power” tools like state media, as well as covered disinformation campaigns.
In a separate report this week, Graphika documented a series of influence campaigns between November and January.
One of them targeted Uniqlo, the Japanese retailer, because he does not use Xinjiang cotton due to concerns about forced labor in the predominantly Muslim region. Another tried to discredit Safeguard Defenders, a Madrid -based human rights organization, using non -authentic accounts on various platforms – including X, Youtube, Facebook, Tiktok, Gettr and Bluesky – to spread false allegations, including those of explicit sexual content.
Laura Harth, Safeguard Defenders’s campaign director, said her researchers faced “a renewed, multilingual attack intended to discredit the work of the organization, threatening, intimidating or defaming some of their team members and trying to sow doubts about their activities” .
c.2025 The New York Times Company
