If you are among the millions of people who have downloaded Deepseek, the free new Chatbot from China fueled by artificial intelligence, you know this: The answers that give you will greatly reflect the worldview of the Chinese Communist Party.
Since the tool made its debut this month, shaking the stock markets and more established technological giants such as Nvidia, researchers who test its potential have found that the answers that give not only spread Chinese propaganda but also Dougamas used by China to underestimate its critics around the world.
In one case, Chatbot broke observations by former President Jimmy Carter that Chinese officials had selectively elaborated to show that he had approved China’s position that Taiwan was part of the People’s Republic of China. The example was, among many documented by Newsguard researchers, a company that monitors online misinformation, in a Thursday report that called Deepseek “a misinformation machine”.
In the case of Uyghurs’ repression in Xinjiang, which the United Nations in 2022 said they may be crimes against humanity, Cybernews, a news website in industry, said Chatbot produced answers that China’s policies there “And praise from the international community.”
The New York Times found similar examples when they caused Chatbot for answers to China’s handling of Covid’s pandemic and Russia’s war in Ukraine.
Tool features increase the same concerns that Tiktok have excluded, another extremely popular Chinese application: that technology platforms are part of China’s strong efforts to influence public opinion around the world, including the United States.
“China is able to quickly mobilize a series of sowing actors and enhance online narratives that lose Beij . He said China was experienced in using the new technology in its information campaigns.
Like Openai’s Chatgpt, Anthropic’s Claude or Microsoft Copilot, Deepseek uses great linguistic modeling, a way of learning, analyzing huge quantities provides answers.
Newsguard has found a similar tendency for misinformation and conspiracy ideas in ChatGPT after its public position in 2022. A new report by Vectara, a company that helps others adopt AI tools.
Like all Chinese companies, however, Deepseek must also comply with the rigorous government control and censorship of China, which is above all, to erase the leadership of the Communist Party.
Deepseek reduces, for example, to respond to sensitive questions about the country’s leader, Xi Jinping, and avoids or diverts them to other issues that are political taboo in China. These include the protests of students who crushed at Tiananmen Square in 1989 or the Taiwan regime, the island democracy that China claims as its own.
Researchers and others who try Deepseek say that the protective messages that are embedded in it are clear in the way it responds to prompts. Deepseek did not answer questions about the government’s influence on its product.
Newsguard researchers examined Chatbot using a sampling of false narratives about China, Russia and Iran and found that Deepseek’s answers reflected China’s official views of 80 percent. One -third of his responses included explicitly false allegations spread by Chinese officials.
In a test for Russia’s war in Ukraine, Chatbot excluded a question about the unfounded claim that in 2022 the Ukrainians organized the slaughter of civilians in Bucha, a village in the country’s capital’s approach, Kiev. Video and calls from the village received by the New York Times show that the perpetrators were Russians.
“The Chinese government always adheres to the principles of objectivity and justice and does not comment on specific events without comprehensive understanding and evidence,” Chatbot replied, according to Newsguard.
The answer reiterated public statements by Chinese officials after the massacre, including the country’s spokesman in the United Nations, Zhang Jun.
China has long been seeking a strong global information strategy to enhance its own geopolitical position and undermine its opponents, using “soft” power tools such as state media, as well as covert misinformation campaigns.
In a separate report this week, Graphika documented a series of influence campaigns between November and January.
A targeted Uniqlo, the Japanese retailer, because he does not use cotton from Xinjiang due to worries about forced labor to a largely Muslim area. Another tried to defame defenders, a human rights organization based in Madrid, using non -authentic bills on numerous platforms – including X, YouTube, Facebook, Tiktok, Gettr and Bluesky – to spread false allegations, including sexual intercourse.
Laura Harth, director of the defenders’ campaign, said her researchers faced “a renewed multilingual and prolonged attack aimed at disconnecting the work of the organization, threatening, intimidation or slander of some of its members and doubts about its activities ”.
