Recently, the Chinese Communist Party has been vigorously promoting the domestic artificial intelligence (AI) platform “DeepSeek”, but its underlying technological basis and political censorship mechanism have raised concerns among the public. In a recent interview with Epoch Times, former Alibaba software test engineer Cai Xiaoli, who is now residing in the United States, revealed from her own experience that the development of Chinese AI technology not only faces structural bottlenecks but is also deeply embedded in the speech control and political propaganda system of the Chinese Communist authorities.
This year on the anniversary of the Tiananmen Square protests, the Chinese AI platform DeepSeek was found to have incorporated speech censorship measures to avoid discussions on the June Fourth incident.
Cai Xiaoli pointed out that platforms like DeepSeek have had strict self-censorship mechanisms built in from the development stage. She mentioned, “When you enter keywords like ‘Chinese Communist Party dictatorship’ or refer to ‘Xi Jinping as a dictator’ on DeepSeek, the system simply won’t respond; no matter how you input, there is no unlocking.” She emphasized that these AI systems have become extensions of the CCP’s propaganda machinery, unable to provide users with diverse perspectives or critical thinking information sources.
She noted that the platform, while showing signs of replicating foreign models in operation, still contains rewritten codes and preset blocking logic in the backend. “Any sensitive topics related to CCP dictatorship, leader controversies, the June Fourth incident, etc., will all be automatically blocked.”
The Chinese startup company “DeepSeek” announced earlier this year that they had trained an AI model at low cost that could rival global leading competitors. However, Cai Xiaoli expressed that DeepSeek is not as powerful as officially proclaimed.
Cai Xiaoli further explained that Chinese AI models, in terms of performance and completeness, fall far short of OpenAI’s ChatGPT, with the major bottleneck arising from a lack of computing resources due to the blockage of advanced chips from the United States.
Furthermore, she indicated that the Chinese algorithm framework and policies themselves have numerous limitations, coupled with closed-off information sources that prevent connections to the knowledge networks of the free world, resulting in a lack of dynamism in the overall innovation environment.
Describing the current state of China’s AI industry as “severely inflated,” Cai Xiaoli criticized many projects as merely serving superficial national propaganda purposes.
When asked why the CCP continues to promote AI achievements, Cai Xiaoli believes it is a “consistent governing tactic” to make the domestic population mistakenly believe that the country is increasingly powerful. She remarked on her past as a ‘Little Pink’ in China, feeling proud under the influence of official propaganda, even going so far as to display the national flag on her face during National Day.
She revealed that she used to harbor negative impressions of Japanese products and people due to official propaganda but later discovered that the Japanese she encountered abroad were polite and respectable. This realization came after she had expanded her exposure to international information.
Cai Xiaoli bluntly stated that the CCP’s “strong nation narrative” fundamentally brainwashes the grassroots population, creating an illusion of progress while actual information censorship intensifies.
In such a scenario, AI platforms like DeepSeek respond only to politically correct terms like “national security” and “national prosperity,” becoming new tools for maintaining political stability. Cai Xiaoli criticized saying, “In a system without freedom of speech, AI will only become the CCP’s mouthpiece.”
Regarding the future development of Chinese AI technology, Cai Xiaoli expressed deep pessimism. She said, “It operates within an authoritarian system with restricted information and speech.”
With a lack of high-end hardware and computing resources, technological development remains limited, compounded by a lack of freedom in the political system. She mentioned that any enterprise operating under such constraints is subject to suppression and restrictions, leading her to have a bleak outlook on its future development.
She even believes that anything encouraged by the CCP should be resisted carefully. “Their encouragement for your involvement is actually for better control over you.”
“I feel that in China, it’s better to ‘lie flat’,” Cai Xiaoli said, stating that under the Communist Party’s system, “efforts yield no returns; you only end up being exploited.”
Reflecting on her transformation from a ‘Little Pink’ to a democratic dissident, Cai Xiaoli shared how she began to awaken after coming into contact with the truth about the Tiananmen Square incident upon scaling the Great Firewall and questioning the information she had received in China.
Apart from the June Fourth events, she also came across authentic information about foreign societal safety and democratic systems, realizing that the foreign reality contradicted the CCP’s claims of “insecurity and daily gunfights.”
After moving to the United States, she officially joined the Chinese Democratic Party and began collaborating in efforts to overthrow the CCP. She stressed the need for continuous action, personal growth, and skill enhancement to drive real change.
In conclusion, Cai Xiaoli’s message to the young people still in China and those who identify as ‘Little Pinks’ is, “The Chinese Communist Party does not equate to China. The CCP is only a temporary ruler, not the entirety of China.”
She emphasized, “A political party shouldn’t have the power to determine the life and death of all Chinese people; such a power arrangement itself is questionable. We Chinese are intelligent and hardworking, but we deserve a free China.”