These days, fake images generated using artificial intelligence, and deepfake videos where celebrities’ heads are superimposed on other people’s bodies, are increasingly more common. It’s hard to tell what’s real and what’s fake, which is problematic.
So how can you avoid being tricked? To help you spot and verify both AI-generated images and deepfake videos and learn about other famous examples, check out this gallery for some helpful tips. Don’t let yourself be fooled!
Tom Cruise hates the Olympics
Russia’s smear campaign against the Paris Olympics made use of AI voice generation bizarrely. This campaign included a documentary made by pro-Russian propagandists attacking the International Olympic Committee (IOC), using an AI-generated version of Tom Cruise’s voice as the narrator.
The creators even added a fake Netflix logo and created a fake New York Times review in an attempt to add legitimacy to the film, but it was quickly spotted as AI manipulation. The attempts to besmirch the Summer Olympics are linked to Russia’s desire to weaken Western support for Ukraine, private experts and Western officials told CNN. The Russian embassy in France called these claims “Russo-phobic.”
Scarlett Johansson vs. ChatGPT
Actress Scarlett Johansson is currently feuding with OpenAI over their adoption of a suspiciously familiar voice for their latest chat-to-voice feature, which they’ve named Sky. Johansson claims that founder Sam Altman approached her in September 2023 to ask if they could use her voice for the tool, stating that it sounded “comforting,” but she declined the offer.
However, when the feature debuted nine months later, she was “shocked, angered, and in disbelief that Mr. Altman would pursue a voice that sounded so eerily similar to mine that my closest friends and news outlets could not tell the difference.” Johansson also said in her statement to the Times that she has hired legal counsel and intends to find out just how Altman developed the new voice of ChatGPT.
Altman has denied using Johansson’s voice but has reportedly paused the feature while he responds to her questions. Altman previously stated that he’s a big fan of the 2013 movie ‘Her,’ in which Johannson voiced the AI character Samantha. He claims that another actress was chosen to voice Sky. The actor’s union SAG-AFTRA has voiced its full support of Johansson as AI becomes a major concern for those in the entertainment industry.
What are AI-generated images?
AI images are imitations of images the program has seen before. It takes existing images found on the internet to create new images, which means the new images aren’t exactly original!
So, how do you check if an image is AI-generated?
First, check out the source of the information you’ve received. Is the source a news organization, a government organization, or the verified account of a celebrity? Or is it just impersonating a legitimate organization or person?
If the image seems to come from a big news story, check the actual news. If something major has happened in the world, chances are international news organizations would know about it first! You should be able to see and/or hear breaking news alerts on TV, radio, and news websites.
It’s also a good idea to look at the image critically. Doing so, you’ll certainly see that it’s an AI-generated image.
They look unreal
Although AI-generated images are usually quite convincing, there’s always a certain quality about the image that looks rather unreal.
Watch for wonky fingers and teeth.Since data sets that train AI systems tend to only capture pieces of hands, the tech often fails to create lifelike human hands, for instance.
A major hallmark of AI images.This can lead to images with stretchy wrists, spindly fingers, or too many digits. Keep an eye out for that. Some AI image generators produce excessively smooth textures, which look too perfect. Also, when it’s an image of a human, the skin can come out looking like plastic with a glossy sheen.
Look for details that don’t fit
Watch out for inconsistencies in the logic of the image itself. For example, clothing fabric that blends across different subjects, or background patterns that repeat perfectly. Apart from the details, make sure you look at the big picture. First, check the lighting and the shadows, as AI often struggles with accurately representing these elements.
For example, shadows should align with the light sources and match the shape of the objects casting them.
If the image you’re looking at contains texts, such as panels, labels, ads, or billboards, take a closer look at them. AI-generated text tends to look pixelated or stretched. Also, if the text doesn’t make sense, is out of context, or contains weird phrases that a human wouldn’t write, the image you’re looking at is likely AI-generated.
Check the title and description
Not everyone will disclose the use of AI when posting images, but for those who do, that information will either be in the title or description section of a post. Another good place to look is in the comments section, where the author might have mentioned it.
When in doubt, always run the image through an AI image detector. Just save the image and pull up an AI image detector service. Free AI image detector services include Illuminarty, Optic AI or Not, and Everypixel Aesthetics.
How to spot deep fake videos
Creating a deepfake video has also become highly popular. So, how do you know if it’s fake?
Examine facial movements carefully.A common warning sign is unnatural eye movements or a lack of eye movement, especially if there’s no blinking. Also, if a person’s face doesn’t show the appropriate emotion that aligns with what they’re saying, it’s likely a deepfake.
Sound quality is a clue.Creators of deepfakes typically focus more on visuals than on audio, so you might notice a difference in the sound or speech. Pay attention to poor lip synchronization, robotic sounds, strange word pronunciation, digital background noise, and even missing audio.
Nonsense words are common in deepfake creation, as they provide a tool for altering or substituting original speech.
A dangerous mix
By blending real and nonsense words, deepfake creators aim to produce convincing audio that imitates the speech patterns of the targeted individual. As a result, viewers may be tricked into believing that a person has said something they haven’t.
Examine the physical posture of the person in the video.If a person turns their head or moves from one frame to another in a jerky way, or if their movements appear distorted or unnatural, be suspicious.
Notice if a person’s body shape doesn’t look natural.Deepfake technology tends to focus on facial features rather than the entire body, making this one of the simpler anomalies to detect.
It’s always important to use your best judgment when seeing a picture or video. Be critical and keep in mind it could be a deepfake or an authentic image.
