AI videos spread Hurricane Melissa misinformation
AI videos spread Hurricane Melissa misinformation
Occurred: October 2025
Page published: October 2025
AI-generated videos falsely depicting extreme scenes of Hurricane Melissa flooded social media, causing confusion and endangering public safety during the Category 5 storm in Jamaica and the Caribbean.
AI-generated videos showing fake images and clips of floods, destruction, sharks in streets, and dolphins in swimming pools spread online ahead of and during Hurricane Melissa hitting Jamaica.
Categorised as a Category 5 storm, Melissa was said to be the most powerful and damaging storm of 2025 and the fake videos, many of which bore watermarks from AI video creation tools like OpenAI’s Sora, overshadowed critical and accurate storm reporting.
This made it difficult for the public to discern trustworthy information and increased confusion during a dangerous natural disaster.
Platforms such as TikTok removed some fake content, but many remained with disclaimers or community notes.
The surge in fake videos was facilitated by the accessibility and power of text-to-video AI technologies, which allow lifelike, convincing videos to be generated and disseminated rapidly online.
Storms naturally attract sensational and emotional content, which encourages viral spread of misinformation. The fake videos sometimes downplayed the storm’s significance or presented fabricated dangerous scenarios, complicating public perception.
These kinds of fake videos hamper real-time emergency communication and public safety efforts, potentially causing harm by delaying protective actions or spreading panic based on false visuals.
More broadly, they highlight the challenge of combating AI-enabled misinformation during crises and the need for improved digital literacy, vigilant fact-checking, and proactive content moderation.
Developer: OpenAI
Country: Jamaica
Sector: Govt - municipal
Purpose: Sow alarm/confusion
Technology: Generative AI; Text-to-video
Issue: Mis/disinformation; Transparency
AIAAIC Repository ID: AIAAIC2090