China has further tightened its censorship and surveillance policies as it works to control the development of artificial intelligence (AI), despite its ambitious pursuit of this technology.
The Chinese Communist Party (CCP) has implemented stricter regulations to ensure that Chinese tech companies comply with its ideological directives.
All AI companies are now required to undergo a government review process that scrutinizes their large language models (LLMs) to ensure they “embody core socialist values,” as reported by the Financial Times last week.
China has long employed its “Great Firewall” to filter internet content, blocking information deemed detrimental to the CCP, such as details about the 1989 Tiananmen Square massacre or memes comparing Chinese President Xi Jinping to Winnie the Pooh.
This firewall is now being extended to the AI domain, not only to restrict the content AI systems generate but also to control the information they process.
China’s Cyberspace Administration of China (CAC) mandates that AI companies like ByteDance, Moonshot, and 01.AI participate in a review process evaluating the effectiveness of their censorship mechanisms in the LLMs they are developing.
Chatbot systems are being designed to identify sensitive keywords and block information related to prohibited topics, often including inquiries about human rights.
These AI systems respond with messages like “try a different question” or “I have not yet learned how to answer this question. I will keep studying to better serve you.”
To prevent excessive blocking, CAC regulations stipulate that LLMs should not reject more than 5% of questions, according to the Financial Times report.
Instead, pre-approved, politically correct responses have been created to answer specific types of questions, though controlling LLM responses remains a challenging task for developers.
China’s relentless efforts to control information dissemination among its own population represent a broader threat, according to AI expert Arthur Herman, senior fellow and director of the Quantum Alliance Initiative with the Hudson Institute, who spoke with The Epoch Times Digital.
“That is the future that China has charted for its own citizens,” Herman said. “This is also how they see… being able to control the world of others.”
Herman highlighted China’s growing influence in the global south, where social media platforms like WeChat have gained significant popularity.
“There will inevitably be a social control, a mind control, element that goes into those programs… and to shape a world that looks more and more like China wants it to look,” he said.
Herman cautioned that these strategies are not limited to authoritarian regimes but pose a risk wherever these platforms are accessible, including the United States.
“They have mastered the art of brainwashing through TikTok,” Herman said. “Chinese engineers have found a way to create a social media platform which is highly addictive, and which is also highly geared towards brainwashing its users to see the world in a certain way and to respond to visual and audio cues in a certain way.”
Herman stated that China’s use of TikTok technologies is merely a “foretaste” of how Beijing might leverage AI applications to manipulate populations beyond its borders.
“China sees AI as a means by which to change people’s minds,” he said. “AI’s ability to enhance those kinds of brainwashing and mind control applications is so powerful…that even when you’re not actually under a surveillance camera, even when you’re not actually listening to or watching government-inspired propaganda… there are other subtler ways in which your mind is being changed and adjusted simply by your interaction with things that are taking place in daily life — which are more and more directed by how the Communist Party wants you to see the world.”