State-Sponsored Amnesia Meets Artificial Intelligence

Typhoon Noul has left a predictable trail of physical destruction in its wake, but Beijing is currently grappling with a secondary, entirely synthetic storm. Across Chinese social media, digital waters are rising. Users are bombarded with hyper-realistic footage of fictitious power outages, non-existent corpses floating through streets, and digital crocodiles supposedly roaming flooded rivers—the latter a creative embellishment on a genuine incident in Hengzhou, where snakes escaped a breeding farm.

This digital chaos is the byproduct of an aggressive state industrial policy colliding with the entrepreneurial spirit of the attention economy. In 2017, Beijing declared artificial intelligence the primary engine of national progress, subsequently pouring vast sums into the sector. Today, the country boasts over 6,000 AI companies.

The technological barrier to entry has plummeted to zero, allowing anyone with a smartphone to manufacture panic at scale. Researchers at the Communist Party's Central Party School acknowledge that prompting an AI tool is now sufficient to generate highly realistic, entirely fabricated content. The motive is rarely ideological, usually driven instead by the simple desire to accumulate followers, farm engagement, and generate traffic for personal gain.

The state’s reaction is predictably authoritarian. On July 23, the Cyberspace Administration of China launched a nationwide crackdown targeting fabricated disaster data, staged scenes, and individuals impersonating officials. Ningbo Public Security Bureau representatives have publicly warned that the internet is not a lawless zone, announcing specialised task forces to monitor short-video platforms. Creators face up to seven years in prison if their digital fictions result in serious real-world consequences. Meanwhile, legal professionals in Shanghai and Washington are already suggesting that the platforms themselves should bear legal responsibility alongside the creators.

Yet, the government is fundamentally battling a crisis of its own making. Authorities implore citizens to rely on accredited sources during emergencies, seemingly blind to the fact that official credibility is entirely bankrupt. The Chinese public resorts to unverified social media rumours precisely because state narratives are routinely manipulated.

Just weeks ago, local journalists were instructed to describe a catastrophic breach at the Liulan Dam merely as an opening. The resulting flood killed 26 people downstream. Similarly, in 2021, officials in Henan province were arrested for concealing 139 deaths. When the state routinely sanitises disaster reporting to protect its image, a trust vacuum inevitably forms.

Detection technology is currently losing the race against generation technology, trapped in an endless game of digital catch-up. Academics advocate for building trusted information infrastructures and early warning systems to verify geographic locations and timestamps. However, technological solutions cannot fix a political problem. A government cannot expect to foster a booming generative AI sector while simultaneously demanding an absolute monopoly on truth from a populace that has learned, through bitter experience, to doubt every official broadcast.

Written by Thorben Thiede thorben.thiede@alpineweekly.com