DE — heise online
Study: AI Videos Significantly Undermine Trust in Moving Images
AI1 min read
A new German study says AI-made videos are making people doubt what they see online. Many viewers have already been fooled, and most want clear rules for labeling AI content.
What happened
The study Video Trends 2026, made by Germany's state media authorities, looked at how people watch videos. Most Germans often see AI-made or heavily changed video clips on the internet. These clips can look very real, and they make people trust video less.
Key numbers
- 93% of people aged 14 and up watch videos every day.
- About 79% say they have already seen AI videos. TikTok and Instagram are the most common places.
- 68% of those who have seen AI videos thought at least one of them was real.
- 81% now doubt the truth of videos more often.
- About 9 in 10 want a general label rule for AI content. 83% want some kinds of deepfakes banned.
Why videos are changing
Social networks and video apps now shape daily life, especially for young people. TV still matters, but many people get their news and opinions from creators. These creators usually focus on entertainment, culture or sports, and most have no journalism training.
Why it matters
Fake-looking videos can fool people, and the signs that used to reveal fakes, like strange facial movements, are getting harder to spot. People who think they are good at spotting fakes are often fooled too. Experts worry that this could change how people form opinions, especially about politics.
What to watch
- Whether lawmakers adopt the labeling and ban rules that most people want.
- How platforms like TikTok and Instagram mark AI-made videos.
- Whether people start checking other sources more often.
Glossary
- Video Trends 2026
- A study by Germany's state media authorities about how people watch videos and what they think of AI clips.
- TikTok
- A short-video app where users watch and post clips.
- A photo and video app owned by Meta, where users share posts and short videos.