China Cracks Down on AI-Generated Rumors About Zhangjiajie

Chinese cyber police have investigated and dealt with two separate cases involving the use of artificial intelligence tools to fabricate rumors about the Zhangjiajie scenic area in Hunan Province, according to Xinhua News. The cases highlight growing concerns over AI-generated misinformation and the authorities’ intensified efforts to combat digital disinformation through the “Clean Network 2026” campaign.

The Two Cases

Case 1: Glass Bridge Collapse Hoax

On April 26, 2026, a 13-second video began circulating on a short-video platform claiming that the Zhangjiajie Grand Canyon Glass Bridge had collapsed. The video showed the bridge platform cracking and tourists falling, accompanied by the caption: “Don’t always think about going out to play, it’s safer at home.”

The video was created by a suspect identified as Ma, a regular internet user who used an AI video generation tool to produce the footage. Ma used a personal photograph he had taken at Zhangjiajie ten years ago as the base image for the AI generation. Within four days, the video had garnered 15,000 likes and over 11,000 shares, causing widespread public panic ahead of the May Day holiday period. Tourists and residents flooded the scenic area with phone calls inquiring about the bridge’s safety.

According to Sina News, which sourced its report from CCTV News, police officer Liu Chuwei of the Cili County Public Security Bureau noted that the video had obvious flaws: “First, the characters’ limbs were very stiff, and there was the characteristic blurring of AI videos, with very heavy splicing痕迹. There was actually an ‘AI-generated’ label at the bottom of the video. This video was fake.”

After the video was reported and removed on April 30, Ma re-uploaded screenshots from the video on May 1. Police traced and apprehended him, and he was placed under criminal compulsory measures for allegedly fabricating and intentionally spreading false information.

Ma later expressed remorse, saying: “I just lacked reverence for the law. I did it for momentary fun, and it caused great harm to society, to myself, and to my family.”

Case 2: Bailong Elevator Hoax

Despite the first case, a second suspect named Zhu was not deterred. On May 23, 2026, Zhu used AI to generate a fabricated article titled “Hunan Zhangjiajie Bailong Elevator Experiences Frequent Failures, Tourists Trapped 100 Meters High for an Hour.” The article falsely described elevator malfunctions and tourists trapped in terrifying scenarios, causing widespread public concern about scenic area safety management.

As CNR reported, cyber police quickly verified that the Bailong Elevator had not experienced any malfunctions or trapped tourists, and the article was entirely fabricated. Zhu, who sought to boost his account’s commercial value, was given administrative detention by local public security authorities.

Broader Context and Legal Framework

Both cases were investigated by the Ministry of Public Security’s Cybersecurity Bureau and are part of the “Clean Network 2026” special campaign targeting online rumors and AI-generated disinformation. On May 26, 2026, the Ministry published five typical cases of cracking down on online rumors, with AI-generated content highlighted as a key category.

As People’s Daily reported, the Ministry of Public Security stated: “For those who intentionally fabricate and spread rumors, public security organs will pursue corresponding legal responsibility according to the degree of harm caused by the rumors.”

This enforcement effort follows a broader campaign by the Cyberspace Administration of China in 2025, which handled over 3,500 AI products and removed more than 960,000 pieces of illegal information.

International Recognition

The glass bridge hoax was catalogued by the OECD AI Incident Monitor on May 16, 2026, marking international recognition of the incident as an AI harm event. The OECD noted that the AI system was “explicitly used to create a fabricated video that falsely showed a dangerous event that did not happen,” leading to social harm by misleading the public.

Analysis and Implications

Police officer Liu Chuwei offered a pointed warning: “AI technology is a tool to make life convenient, but it is absolutely not a tool for us to casually spread rumors, seek attention, or earn traffic. The internet is not a lawless place. Never cross the legal底线 for a bit of traffic.”

The two cases reveal an escalating pattern of AI misuse for disinformation. The first suspect used AI to generate a video from a personal photo, while the second used AI to produce a full fabricated article. Both incidents were timed ahead of major travel periods, suggesting deliberate targeting of high-traffic tourism seasons to maximize impact.

What’s Next

As China’s “Clean Network 2026” campaign continues, authorities are expected to maintain pressure on AI-generated disinformation. The cases serve as both a deterrent and an educational tool about detecting AI-generated content. For the tourism industry, the incidents underscore the economic threat posed by AI-powered rumors, particularly for high-profile destinations like Zhangjiajie that rely heavily on visitor confidence.

The different legal outcomes — criminal compulsory measures for the video hoax that caused direct public panic, versus administrative detention for the article fabrication — reflect a calibrated enforcement approach based on the severity of harm caused.