GLOBAL - As Japan confronts an unprecedented wave of bear attacks resulting in a record number of fatalities, a parallel threat is emerging online: a flood of artificial intelligence-generated videos depicting fabricated encounters with the animals. This digital misinformation is stoking public anxiety and raising concerns among experts that it could dangerously mislead the public on how to respond to a genuine threat, according to multiple reports published on Thursday, 13 November 2025, EAT.

Since April 2025, Japan has recorded at least 13 deaths from bear attacks, more than double the previous high, with over 100 people injured, as confirmed by Japan's Ministry of the Environment. This sharp increase in human-bear conflict is attributed to several factors, including food shortages for the bears—linked to poor acorn and beechnut crops—and the depopulation of rural areas, which has blurred the natural boundaries between wildlife habitats and human settlements.

Amid this heightened state of alert, social media platforms like TikTok have been inundated with hyper-realistic but entirely fake videos. An investigation by Japan's Yomiuri Shimbun newspaper found that approximately 60% of 100 examined clips related to bears were created using OpenAI's Sora video generator. These videos range from the dramatic—showing bears destroying solar panels or snatching pet dogs—to the deceptively mundane, such as a news report on local sightings or an elderly woman feeding fruit to a bear. Some clips even depict dangerous and incorrect responses to an encounter, such as a schoolgirl fighting off a bear.

The Dual Threat: Real Bears and Fake News

Japan is home to two bear species: the Asiatic black bear, found on the main island of Honshu, and the larger Ussuri brown bear in Hokkaido. Adult black bears can weigh up to 130kg, while Ussuri brown bears can reach 400kg, and both can easily outrun a human. The viral AI videos often depict behaviors and scenarios that wildlife experts warn could lead to tragic outcomes if emulated. Shinsuke Koike, a professor at Tokyo University of Agriculture and Technology, told the Yomiuri Shimbun that feeding a bear is “extremely dangerous” as it could cause them to lose their fear of humans.

The Japan Fact-Check Center has noted a significant increase in these fake videos, which often gain credibility by referencing real locations where bear sightings have been reported. Daisuke Furuta, the center's chief editor, has urged the public to be aware of the risk of deception. In one instance, a fake news report of a bear capture in Noshiro, Akita Prefecture, prompted city officials to issue a denial, stating such misinformation could cause public unrest and disrupt official work.

The spread of this content complicates the Japanese government's public safety messaging. Authorities have been advising residents in affected areas to carry bells, avoid hiking alone, and, if confronted, to back away slowly without turning their back. If attacked, the official advice is to lie face down and protect the head and neck. The AI videos, which show people fighting back or interacting casually with bears, directly contradict this expert guidance and could undermine public trust in official sources.

Broader Implications for a Digital World

The situation in Japan serves as a stark case study of the broader global challenge posed by AI-generated misinformation, particularly when it intersects with public safety and wildlife conservation. Experts in conservation biology have warned that misinformation can harm conservation efforts by creating false narratives and promoting harmful human-wildlife interactions. The realistic nature of modern AI tools makes it increasingly difficult for the public to distinguish fact from fiction, a problem that extends far beyond wildlife encounters to political and social stability.

In response to the growing threat of AI-driven disinformation, the Japanese government has been enhancing its countermeasures. In May 2025, Japan's parliament passed its first law to address AI misuse, empowering the government to investigate and issue warnings, though without penalties. A cross-governmental unit was also established in 2024 to analyze and counter foreign disinformation campaigns. However, the rapid, decentralized spread of content on social media presents a formidable challenge that technology and legislation are still struggling to address effectively.

As Japanese authorities continue to manage the on-the-ground crisis of increased bear encounters—deploying soldiers to assist hunters and considering revisions to hunting rules—they must now also contend with a viral wave of digital falsehoods that threatens to make a dangerous situation even worse. The convergence of these two crises underscores a new and complex reality for governments and citizens worldwide.