How to Spot AI-Generated Images and Text
AI-generated pictures and text are now common online. There is no perfect test, and the tools keep improving, but a few habits can help you decide how much to trust what you see.
Clues in images
- Hands, teeth, jewelry or text in the background that look warped or inconsistent
- Lighting or shadows that do not match across the scene
- Skin or objects that look unusually smooth
- Details that repeat in odd ways, such as identical windows or faces
Absence of these clues does not prove an image is real, and their presence does not prove it is fake.
Clues in writing
- Smooth but generic phrasing with few concrete details
- Confident claims with no source
- Repeated sentence structures
- Facts that do not hold up when you check them
Check the source, not just the content
A reverse image search can show where a picture first appeared. For text, look for an author, a date and links to original sources. Trustworthy sites usually tell you who is behind the content.
Slow down before you share
Surprising or emotional posts spread fast. Pausing for thirty seconds to check the source is often enough to avoid passing along something false.
Detection tools exist, but their results can be unreliable, so use them as one clue among several rather than the final word.
