In today’s online world, telling what’s real from what’s fake can be tricky. As artificial intelligence (AI) gets better, we’re seeing more fake images and videos. Sometimes, it’s fun to see famous people in places they’ve never been to. But there’s also a bad side to it.
These fakes, known as “deepfakes,” are made using tools that anyone can use, not just computer experts. While it might seem funny to see a celebrity in a weird situation, these fakes can be used to scam people, steal their identities, or even mess with important events like elections.
That’s why it’s important we all learn how to spot these fakes.
How to Spot Fake Images
To tell if an image is a deepfake, start by looking at the small stuff. In older fake photos, mistakes were easy to spot—like a person having three hands or glasses that looked odd. Today’s fakes are trickier because the tech is better. But their creators still mess up sometimes.
Check the Image for Inconsistencies
Look for inconsistencies in lighting and shadows, image quality, skin texture, and blemishes in the photo. For instance, if the light on someone’s face doesn’t match the light in the rest of the picture, or if shadows are all over the place, that’s a red flag. Also, check if the main thing in the photo looks really clear and real, but the stuff around it or in the back doesn’t seem right. This mismatch can be a big hint that the image has been messed with.
Examine the Credibility of What You See
When you’re scrolling through photos online and stop because something looks weird, think if it makes sense. Often, we see pictures of famous people doing unusual things that just don’t fit with what we know about them. If you see something like a well-known politician wearing the latest in fashion, it’s time to question what you’re looking at.
Remember, big events that get lots of attention like the Met Gala have heaps of photos taken by pros. Use those photos as a guide to check if what you’re seeing is real. A good clue may be something small but off like how everyone talked about a Katy Perry photo from the Met Gala. The photo seemed fine until people noticed the carpet from a previous event. If the background or situation in a photo looks strange, there’s a good chance it’s not real.
Check the Source of the Image
It’s essential to be a critical thinker and consider where the image came from to determine its authenticity. Knowing the source can give clues about the image’s purpose and origin. Is it from a satirical website or page? Then the image is likely intended to be humorous.
Reputable sources like established news organizations, professional photographers’ websites, or verified social media accounts of the people in the images are more likely to be genuine.
Ultimately, be wary if an image does not have context or attribution.
Do a Reverse Search
Use a search engine like Google Images to see if the image appears elsewhere online. If it’s a deepfake, it might have been used in other contexts. If you find similar images, check for manipulations like a person that may have been replaced in the fake image.
Use AI Tools to Find AI Fakes
It may sound like a plot twist in a movie, but we can use AI to spot fake images created by other AI systems. Companies like OpenAI, Microsoft, and Intel are involved in this endeavor. For example, OpenAI made a tool that checks if images were made using their DALL-E 3 system. Microsoft’s tool looks over photos or videos and gives a score on whether it thinks they’ve been tampered with. Intel’s FakeCatcher goes deep, checking the very pixels of an image to see if it’s the real deal or a fake.
Similar to an AI checker that checks written text, AI-based tools can help us spot fake images, too. However, not everyone can use these tools yet. They’re still being tested and have yet to be made available to everyone. Why? Well, their makers don’t want the bad guys who make deepfakes learn how to recognize what’s real from what’s fake.
These essential tips to spot AI-generated deepfake images can help you protect against online threats.
The Evolving Challenge of Identifying Deepfakes
Telling what’s real from what’s fake is like a race where the finish line keeps moving. AI is getting smarter by using a ton of stuff it finds online to make images, audios, and videos that look and sound very real. What works for finding fakes today may not work tomorrow.
Experts are pretty worried about this situation. They say if we think we can spot every fake, we may end up fooling ourselves. It’s not so simple to always tell what’s real from what’s not, even for people who are really good at it.
It’s a big ask to always wonder if what you’re looking at is the real deal. But staying on top of things and knowing about the challenge is a good place to start. It’s important to keep an open mind and be ready to change how you spot fakes.
Sources
- https://www.unicef.org.au/unicef-youth/staying-safe-online/how-to-spot-fake-news-and-images
- https://www.bbc.com/future/article/20240311-how-to-spot-a-manipulated-image
- https://guides.library.illinoisstate.edu/evaluating/images
