Five lessons from the Nepal floods, as AI worsens audio-visual truth in climate disasters
Following a devastating glacial collapse and flood in Nepal, we analyzed how artificial intelligence impacts the reliability of information during climate disasters. The report highlights that current AI detection tools are often unreliable when analyzing environmental scenes rather than human faces, and it advocates for standardized provenance labeling and better support for local fact-checkers to ensure that manipulated content does not obstruct essential humanitarian efforts.
| Collection(s) | All resources, Technology, Threats & Opportunities: Research & Analysis |
|---|---|
| Region/Country | Asia |
| Topic(s) | Deepfakes, Open Source Investigations |
| Type | Report |
| Language | English |
