U.S. House Committee reports 85% of DeepSeek AI chatbot responses align with Chinese Communist Party narrative.

According to Taipei Times a recent report by the U.S. House Select Committee on the Chinese Communist Party shows that approximately 85% of the responses from the Chinese AI chatbot DeepSeek are filtered or altered to reflect the Chinese Communist Party’s (CCP) narrative. A Hangzhou-based startup linked to the Chinese Communist Party launched the chatbot called DeepSeek on January 20 which has been accused of acting as a “digital enforcer” to suppress dissent and skew information on topics such as democracy, Taiwan, Hong Kong, and Chinese human rights abuses to comply with Chinese laws.

According to the committee, DeepSeek is a “profound threat” to the United States security as it has distorted the global discourse. It has been established that Liang Wenfeng, the co-founder of High-Flyer hedge fund, controls the startup, and the latter employs automated filtering and biased algorithms to ensure that outputs conform to the CCP ideology. The report states that the manipulation of the information erases the opposing views and this raises a lot of concern about the impact of the chatbot to international users.

The operations of DeepSeek also pose a threat because of the data practices employed in the company. The committee established that the backend of the chatbot is linked to China Mobile—a U.S.-designated Chinese military entity—and that it sends U.S. user data to the CCP. It is noted that the data gathered from millions of American users is considered as a “high-value open-source intelligence asset” to Beijing, which poses a great threat to privacy and security. The report further states that DeepSeek probably practised unauthorized model distillation, which entails the process of copying the logic of other AI models and possibly violating the intellectual property rights.

The committee established that DeepSeek employees used nicknames and a multi-jurisdictional banking structure to penetrate U.S.-based AI platforms, including OpenAI. The committee received confirmation of these findings from OpenAI in its statement to the committee that raised further concerns about the clandestine activities. Furthermore, there is also controversy about the acquisition of tens of thousands of Nvidia chips by DeepSeek as the latter is prohibited from importing such chips from the US. According to Taipei Times, the U.S. Department of Commerce is investigating whether DeepSeek has purchased these advanced chips through Singapore, a country that has less strict export regulations.

The report highlights the existing competition between the US and China in the tech industry, especially in the areas of artificial intelligence and semiconductors. DeepSeek’s rapid growth, which has been attributed to the low-cost AI model R1, has energized China’s technology industry, but its relationship with the CCP and its alleged wrongdoing have raised concerns. The committee’s results show the dangers of unregulated use of artificial intelligence, particularly when the platforms serve as the means of transmitting state information. For the U.S., the challenge is in the area of promoting innovation while at the same time preventing foreign interference and data exploitation.

The ongoing probe by the Commerce Department may change the course of U.S. policy on AI and export controls as a result of DeepSeek’s actions. DeepSeek’s operations are a clear indication of the CCP’s plan to use technology to exert its influence across the globe and the report urged increased caution to protect national security and user privacy. This investigation sets the stage for a very important conversation about the relationship between AI, geopolitics, and data sovereignty.