Several Chinese cybersecurity companies have recently announced the integration of the artificial intelligence large language model DeepSeek into their services. These companies include Tianrongxin, Qi An Xin, and NetEase Yidun, all of which have long provided cyber censorship services to the government. They claim that DeepSeek enhances content moderation capabilities, enabling more precise identification of homophones, metaphors, and other hidden risks while improving threat assessment and information monitoring.

Experts warn that DeepSeek’s advanced language processing capabilities will further bolster the Chinese government’s ability to censor speech and control society. Benny Casa, a researcher at the Swiss Federal Institute of Technology in Zurich, pointed out that this tool can rapidly scan and classify vast amounts of data, monitoring political, military, and social discourse to provide authorities with faster response times. Additionally, large language models can also be used in cyberattacks. For example, a recent OpenAI report indicated that suspected China-based accounts were using ChatGPT to attack dissidents on social media.

China has already established itself as a global leader in surveillance technology, and the widespread adoption of DeepSeek will further consolidate this advantage. Government agencies, state-owned enterprises, banks, hospitals, and major internet companies are all integrating DeepSeek, reflecting Beijing’s strategy to localize technology development. Experts believe that China is accelerating its efforts to reduce reliance on Western technology while deepening the application of AI in surveillance, further tightening control over domestic information flows.

(Source: Voice of America)