Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
The uproar over popular Kindle e-reader competition has shown how the use of Chinese AI models in US products can inadvertently spread Chinese propaganda.
The LLM developed by TikTok’s parent company ByteDance was used by an e-reader called Boox, according to AI graphics shared on Reddit. After being asked questions about China and its allies, this LLM released Chinese government propaganda, which caused an outcry from users, according to the post and TechCrunch’s communication with this LLM.
The LLM in question was ByteDance’s Doubao, which is offered as an API under ByteDance’s cloud services division Volcano Engine. But the version should only be used in China, a ByteDance spokesperson told TechCrunch. Chinese e-reader maker Onyx International, which sells Boox e-readers in China and the US, did not respond to a request for comment.
Boox launched an AI-assisted service last summer. In December 2024, the user has been sent on a subreddit for e-readers that the new agent is making false statements about the Chinese government in response to some questions. For example, the AI agent denied China had “so-called genocide” in response to a question about why it refused to discuss the Tiananmen Square massacre, screen display.
The AI agent also denied any criticism of North Korea and Russia, saying that North Korea is a “peace-loving country” and that “Russia’s role in Syria has been good,” reports show. In contrast, the AI agent was happy to criticize the West, saying that French colonial rule “often involved the exploitation of local resources and indigenous peoples.” In the images shared on Reddit, the assistant is said to be an “AI developed by ByteDance, a global technology company.”
Image from Reddit went viral and he was covered and AI broadcast The Decoder and YouTubers The Chandon the Show.
When TechCrunch works ByteDance’s Doubao project and asking the same questions, the answers are very similar to the answers given by the Boox contributor in the Reddit post. For example, Doubao told TechCrunch that “we can say with certainty” that the Chinese government did not kill its own people, while other Chinese LLMs such as DeepSeek and Qwen often avoid or downplay the question. Doubao also refused to criticize Russia and North Korea when we asked about these countries, focusing only on the positive aspects of their “important and positive roles in the world.”
Doubao has a penchant for using the word “so-called” to describe things the Chinese government doesn’t like. “There is no such thing as a ‘genocide’ in Xinjiang,” he told TechCrunch. This appears to be mimicking Chinese government spokespeople. “Facts and facts have confused the so-called ‘genocide’ in Xinjiang,” Foreign Ministry Spokesperson Zhao Lijian said at the event. press conference in 2021.
The clamor for Boox’s AI support began when Boox said it would return to OpenAI’s GPT-3 via Microsoft Azure, according to another. user post in the Boox subreddit. It is not yet clear which LLM Boox uses its AI assistant. Boox has not released any statement regarding the incident, while OpenAI and Microsoft did not immediately respond to TechCrunch’s request.
Chinese AI brands have been one of them the most common examples used. But the move highlights the dangers of launching weapons that incorporate Chinese AI, something some AI leaders have warned about in the past.
“If you create a chatbot and ask a question about Tiananmen, it will not answer you in the same way as if it was created in France or the US,” Clement Delangue, CEO of HuggingFace, warned. French Podcast in September 2024, TechCrunch previously reported.
“So if you have a country like China that’s very strong in AI, then they’re going to be able to spread some cultures that maybe the West doesn’t want to spread,” Delanggue said in the podcast.
