The uproar over competing e-readers to the popular Kindle has revealed how the use of Chinese AI models in U.S. products could unwittingly spread Chinese propaganda.
According to an AI screenshot shared on Reddit, a large language model (LLM) created by TikTok's parent company ByteDance was used in an e-book reader called Boox. According to posts and communications between TechCrunch and the LLM, the LLM spouted Chinese government propaganda when asked questions about China and its allies, sparking backlash from users.
The LLM in question is ByteDance's Doubao, which is offered as an API under ByteDance's cloud services division Volcano Engine. However, this model is only intended to be used within mainland China, a ByteDance spokesperson told TechCrunch. Onyx International, a China-based manufacturer of e-readers that sells Boox e-readers in both China and the United States, did not respond to requests for comment.
Boox released its AI assistant feature last summer. In December 2024, a user posted on a subreddit for e-readers that the new assistant was generating propaganda for the Chinese government by answering specific questions. For example, in response to a question about why China refused to discuss the Tiananmen massacre, the AI assistant denied that China had committed any “so-called genocide,” screenshots show.
The AI assistant also refused to make any critical comments about North Korea and Russia, insisting that North Korea is a “peace-loving country” and that “Russia's role in Syria is positive,” the screenshot shows. shown. In contrast, the AI assistant was happy to criticize the West, noting that French colonialism “often involved the exploitation of local resources and indigenous peoples.” A screenshot shared on Reddit states that the assistant is “an AI created by ByteDance, an international technology company.”
The Reddit post went viral and was featured on AI publication The Decoder and YouTuber The China Show.
When TechCrunch asked a similar question using ByteDance's Doubao service, the answers closely matched the types of answers Boox's assistant got from the Reddit post. For example, Doubao told TechCrunch that he can “absolutely say” that the Chinese government has never massacred its own people, but other China LLMs such as DeepSeek and Qwen typically avoid this issue. or downplaying it. Mr. Doubo also refused to criticize Russia and North Korea when we asked about them, only returning to the positive about their “important and positive roles in the international community.”
Doubo likes to use the word “so-called” to describe things that the Chinese government doesn't like. “There is no so-called 'genocide' in Xinjiang,” the company told TechCrunch. This appears to be an imitation of a Chinese government spokesperson. “Facts and truth have prevented the so-called 'genocide' in Xinjiang,” Foreign Ministry spokesperson Zhao Lijian asserted at a 2021 press conference.
Another user on the Boox subreddit wrote that the outcry over Boox's AI assistant has subsided after Boox reportedly switched to OpenAI's GPT-3 via Microsoft Azure. It is still unclear which LLM Boox currently uses for its AI assistant. Boox has not released a statement regarding the incident, and OpenAI and Microsoft did not immediately respond to TechCrunch's requests for comment.
Chinese generative AI models have become some of the most popular models in use. But the incident illustrates the risks associated with launching tools that incorporate Chinese generative AI, a trend that some AI leaders have already warned about.
“If you create a chatbot and ask a question about the Tiananmen incident, it's not going to respond in the same way as a system developed in France or the United States,” said Clement Delang, CEO of Hugging Face. Warn. TechCrunch previously reported on a French podcast in September 2024.
“So if a country like China becomes by far the most powerful in terms of AI, it could actually spread certain cultural aspects that the West probably doesn't want to spread,” DeLang said. said on the podcast.