AI is a content machine that can be used for good and bad. We share tips on how to spot AI generated content designed to pull the wool over your eyes.
AI is a powerful content creation tool, allowing us to create quality text, image and video content faster than ever. While it can make brainstorming and content production more efficient, the downside is that AI makes it harder to tell what’s authentic and what’s not. And that’s where the danger lies, leaving us more gullible to scams, fake news and more.
According to Capitol University, the rise of AI has also contributed to increased fraud.
Not all of us have access to AI detector tools, and by the time you get to the bottom of this piece, you’ll be an almost qualified AI detective.
How to spot AI-generated videos
I’m sure you’ve come across a video of an old woman outrunning a bear or a YouTube exclusive on a random channel detailing (false) information on a scandal with a monotone voiceover and looping images. While they are somehow intriguing, a lot of the time, they are AI generated.
When it comes to videos of people, AI hasn’t mastered human movement and often glitches in subtle ways: excessive blinking, showing strange shadows, or unnatural repeated movements that don’t quite make sense.
If the video has seamlessly perfect voiceover with no natural breath sounds or pauses, it’s probably AI-generated.
Plus, AI struggles with finer details like hands and feet, so AI generated videos tend to not show these but if they do, they are usually suspicious.
Image sourced from The Conversation
How to spot AI-generated images
Remember the viral image of the late Pope Francis I wearing a white Balenciaga puffer coat? Millions of people believed that image was real and it generated lots of engagement and hoo-hah for obvious reasons. The Pope – in a Balenciaga puffer? The image had a lot of shock value and that’s a good reason to question the authenticity of any information.
AI generated images are often too polished and overly perfect. While AI can generate expressive faces, the faces often appear robotic, there’s no soul or depth behind them. The hands often look like cutouts and are disproportionate, and the skin and hairline are overly smooth.
Can you tell the difference between real or fake images? Take the test
How to spot AI-generated text
The tone is usually the first clue. AI generated text lacks a true humanistic and personal tone. It’s often too polished, it never makes errors, even as poetic licence and is also very generic. It says all the right things without saying anything particularly original.
Protecting yourself from misleading AI-generated content
With all that said, perfection seems to be AI’s kryptonite. Human error encompasses the kinks that make us human: the umming between words and thoughts, the rough edges around our hair no matter how much we do it up, the clumsy reflexes that make us make typos and more. These are the kinks that AI cannot mimic.
AI isn’t inherently bad. It’s a tool. The issue is transparency. When people use AI to mislead, impersonate, or create fake narratives, that’s when it becomes a problem.
The best way to protect yourself is to stay curious. Question what you see. And if something feels off, it probably is. In a world where content can be generated in seconds, critical thinking is your biggest advantage.
To get more insights on how to spot AI-generated content designed to fool you, listen to our podcast on AI fraud.
