Muslim community in Japan shaken by suspected arson attacks: ‘everyone is scared’

A series of suspected arson attacks has terrorized the Muslim community in Ebetsu, Hokkaido, Japan, in a disturbing case of online disinformation spilling into real-world violence. The incidents, which include three fires in just over two weeks targeting a Pakistani-owned vehicle repair business and the local Ebetsu Masjid mosque, are being investigated by police. Community leaders report the violence was preceded by months of social media agitation where AI-generated posts and synthetic media falsely blamed Muslims for local construction violations, creating a hostile atmosphere that culminated in physical attacks. One man was injured jumping from a window to escape a blaze, and the 138-square-meter mosque was severely damaged.

The attacks underscore the dangerous, tangible consequences of AI-powered disinformation campaigns, particularly when targeting minority groups. The community of roughly 700 Muslims, many long-term residents like mosque chairman Mohammed Imran, now lives in fear, with the mosque—typically left unlocked for prayer—becoming a symbol of their vulnerability. This event serves as a critical case study in how generative AI tools can be weaponized to fabricate narratives, inflame local tensions, and directly incite violence, challenging law enforcement and platform moderators to identify and mitigate synthetic harmful content before it escalates beyond the digital realm.