A deepfake video fabricating an online conversation between prominent Hong Kong activists has become the first known exercise in the next level of AI-enabled influence operations. Presumably concocted by the Chinese government or a group that serves it, the video heralds a more tailored and comprehensive approach of psychological warfare.
As China integrates deepfake technology into its influence-operations playbook, psychological operations could soon use the technology as a tool of emotional and cognitive manipulation.
The deepfake video purported to show a leaked secret conversation between exiled Hong Kong activists discussing their concerns over the possibility of being extradited from Britain to China as a result of Britain’s plan to reinstate an extradition deal with Hong Kong. The video was first published by the Facebook page Yellowbrainclown 黃腦膠戰 on 26 July. Within two hours of the original post appearing, 22 other accounts shared the video to 17 Facebook groups.
Some accounts showed signs of inauthentic behaviour, appearing to post across different Facebook groups within the same minute. For example, an account under the name of ‘Pocky Miu’ shared the video to two community groups and two pro-Beijing groups at 11:05 am Hong Kong time. Additionally, many of the accounts had characteristics typical of spam or proxy accounts. Such characteristics include the absence of personal information, use of scenery images for profile pictures, and absence of other public posts except for pro-China content. Notably, 12 accounts had only lately joined their Facebook groups, between January and June 2025.
While the operation was likely meant to target Hongkongers—especially the pro-democracy circle—the deepfake video did not gain much traction. It stayed mostly in the echo chamber of the pro-Beijing camp.
Its importance, rather, is as a harbinger of future AI-enabled influence operations.
It shows that psychological warfare remains at the heart of China’s hybrid warfare strategies. Psychological warfare targets adversaries’ military personnel, policymakers and relevant key figures. In the case of Hong Kong, authorities in Beijing and Hong Kong see prominent exiled activists—especially those who have significant international influence and are capable of mobilising diaspora communities, engaging in lobbying or leading movements—as targets for psychological operations.
Such influence operations serve multiple purposes. By fabricating a deepfake conversation of three prominent exiled activists, the operation was likely meant to foment distrust of and disappointment in the British government among Hongkongers by asserting that they are regarded as expendable pawns, no longer holding strategic value to Western countries. It likely aimed to incite fear and chaos within the Hong Kong advocacy circle while amplifying scepticism between Hongkongers and the British government. This could ultimately undermine their trust in liberal democracy and discourage the social movement and international advocacy.
The deepfake video itself was poorly made: words and lips did not sync, the synthetic voice sounded odd and unnatural, and apparent instances of lagging saw facial expressions freeze and repeat. Yet the nature of the content, narrative and tactics targeted the vulnerabilities of the Hong Kong diaspora. The possibility that London may allow case-by-case extradition of Hongkongers had already sparked widespread fear and backlash amongst those who had sought refuge in Britain after pro-democracy protests in 2019. If such influence operations were orchestrated by more capable and malign state actors, the effect could be far more severe.
In recent years, Chinese influence operations and cognitive warfare have increasingly focused on using controversial issues in liberal democracies to exacerbate division. With deepfake technology integrated into the Chinese influence operation playbook, the era of sophisticated cross-domain operation has emerged. The deepfake conversation presented in the video marked this increasingly tailored and comprehensive approach to psychological warfare.
Nevertheless, the video is likely just an early test for China’s future operations. The operation showed no signs of certain tactics, techniques and procedures common in Chinese information operations on Hong Kong-related issues, suggesting that it was merely testing the waters. Examples of typical activities include cross-posting and large-scale dissemination through multitudinous hired commenters (nicknamed Beijing’s 50 Cent Army) and use of political spam networks, as well as amplification by major key opinion leaders and grey media in Hong Kong.
The greatest potential of weaponised deepfakes lies in maximising psychological impact through the presentation of supposed visual evidence, sparking fear and enhancing the persuasiveness of an allegation or conspiracy theory. We are likely to see more large-scale AI-enabled psychological operations leveraging deepfakes, not only for disinformation, but for emotional and cognitive manipulation through the creation of persuasive narratives.
