Example *1
“Meteor-like attack on Tel Aviv”
A dramatic war clip can feel urgent and believable, which is exactly why it spreads fast before people check it.
Preview Page · Not in Top Menu
DeepFake, Verification, and AI Literacy
This public-facing page is designed as a guided entry point rather than a research abstract. Different audiences can start from different paths, and the example images below are tied to real public incidents and their original sources. The purpose is to help people verify suspicious media without forcing them to read a full technical paper first.
Choose a Learning Path
General Public
Traditional Chinese course for adults, journalists, workshop participants, and the general public.
Kids Deck
Story-based teaching slides for parents, teachers, and classroom use.
Grades 3–4
Lower-threshold game route with simpler tasks and lighter wording.
Grades 5–6
Higher-grade version with stronger emphasis on source tracing and tools.
Detection Demo
Hands-on detector entry point for people who want to test suspicious media directly.
Real Cases Used in Teaching
The cards below do not use random stock visuals. Each image points back to the original report or article for that specific event.
Example *2
“20 years ago Lai speech” fake clip
Not every false video is fully AI-generated. Some are edited, reframed, and circulated as if they were real historical footage.
Example *3
“90-year-old duet” AI video
Warm and emotional clips are often shared without checking. That makes them useful teaching cases for public verification.
Example *4
Fake audio of Taiwan’s president
Verification is not only about faces. Audio rumors travel quickly too, especially once clips are reposted across platforms.
Reference *5
READr Deepfake event archive
Some people need one case; others need the wider pattern. This archive is useful for seeing how Deepfake events accumulate over time.
Guide *6
FactLink newsroom guideline
For people who want a practical reference rather than only a case story, this guideline explains how AI-generated image verification can be handled in a newsroom workflow.
Video Previews from the Teaching Site
What the Public Version Teaches
Pause First
Do not forward a surprising clip just because it feels urgent or emotionally powerful.
Check the Source
Find the original post, publication, or fact check before trusting a screenshot or repost.
Use Tools
Reverse search and comparison tools are practical skills, not only technical tricks for experts.
Understand Limits
AI detectors help, but the point is better verification with evidence, not blind trust in one score.
Reference Notes
- TFC, “網傳『特拉維夫被流星雨式空襲』影像，極可能為AI生成”
- TFC, “20年前賴清德演講是虛構變造的影片”
- TFC, “九旬老夫婦二重奏是虛構的AI生成影片”
- RFA / Asia Fact Check Lab, leaked audio case
- READr deepfake event archive
- FactLink, AI for Trust newsroom guideline
All external images on this page are attached to the corresponding original report or article above. They are included as cited event references so readers can immediately trace the source context.