Chinese AI systems are heavily censoring the spiritual group Falun Gong, according to a new report from the Australian Strategic Policy Institute.
The study found that major Chinese-developed models produce zero responses in Chinese.
They also produce far fewer English responses on Falun Gong, a spritual practice persecuted by the Chinese Communist Party (CCP) since 1999, than any other sensitive topic.
The report shows these systems not only refuse to answer but also distort narratives.
And they echo the CCP propaganda by framing Falun Gong as a public threat.
Falun Gong is a peaceful meditation practice based on the values of truthfulness, compassion, and forbearance, with millions of practitioners in more than a hundred countries.
Researchers warn this pattern demonstrates how AI is being weaponised to export censorship and political influence beyond China’s borders.
China is already the world’s largest exporter of AI surveillance technologies.
Experts said democratic societies must adopt safeguards to protect free expression and prevent authoritarian messaging from spreading through AI tools globally.
