AI (Artificial Intelligence) is no longer just a buzzword, it’s everywhere, that’s why spotting AI is so important! From social media to the arts and media! With it becoming so mainstream, the line between what’s real and what’s not becomes blurred.
Unfortunately, you can’t check every piece of media with an AI content detector. So, improving your AI spotting skills is a must! Sometimes, content is conveniently marked as AI-created. Other times, it’s simply not. Avoid being misled, start spotting AI-generated content today.
AI-generated images
AI-generated images often look a bit too perfect! The lighting might seem too even, and the contents of the image can appear glossy and artificial. AI struggles with the finer details, so always make sure to check for distorted imagery in the background of an image and look out for garbled text and numbers. Also, if there are persons present in an image, check that they don’t have extra or misplaced hands and limbs. This is a dead giveaway!
For detailed instructions on spotting AI-generated images, make sure to check out our dedicated blog. You’ll be a pro AI content detector in no time!
Spotting AI-generated videos
The first thing you should be thinking to yourself is, “Does this scenario seem believable?” If a video’s contents are unbelievable, or too good to be true, stay suspicious and look for other signs. Like AI-generated images, an obvious sign that a video is AI-generated is if the quality of the video looks too perfect, glossy and artificial.
We’d also suggest looking out for discontinuity when spotting AI content. For example, an item of clothing might change colour across different clips or different elements may blend into each other.
Unnatural body posture or facial movements is another giveaway. If the mouth or chin area looks blurred and the audio doesn’t sync with the mouth, treat the video’s origins with suspicion!
With Google’s release of Veo 3, their latest AI video model, AI-generated videos have become a lot harder to spot! It’s worth mentioning, AI-generated videos don’t always look polished or high quality.
Many people creating and posting AI-generated videos are now using poorer quality clips to make them more believable. Passing them off as security camera footage or clips filmed on a mobile phone.
Audio quality is currently another way that you can spot AI videos. Voices often sound metallic and tinny. Read our next section to learn more. Keep in mind, updates will likely fix these flaws very soon. AI is rapidly evolving and will likely not be distinguishable from non-AI-generated content in what could be a matter of months.
AI-generated audio
AI has provided the ability to copy a person’s voice and pretend to be someone else. This also makes it a great tool for scammers who can imitate a loved one or someone you know. Unfortunately, you won’t be able to use an AI content detector for a phone call.
So, if you get a call or a voice message that seems suspicious, listen closely. Is there oddly no background noise at all? Or is there background noise that sounds unnatural, such as static or crackling noises?
Oddities like this should be an indication that something is not legitimate. If you pay attention to the speaker’s voice, does their tone sound robotic, flat or unnatural? Context clues are also important to pay attention to. It’s strange to get a call from your high school friend saying, “Hello, I’m having trouble remembering the name of our history teacher, could you help me out?”
An unexpected call out of the blue like this could be an attempt to gather personal information to access accounts. Another common tactic is using the voice of a loved one in distress. Err on the side of caution; you could be very likely listening to AI-generated audio.
Make a safeword with your loved ones.
We suggest families create a “safeword” that only they know. This can be used to verify their identity to each other. This method has even been recommended by the FBI.
According to the FBI “Criminals exploit generative artificial intelligence (AI) to commit fraud on a larger scale, which increases the believability of their schemes. Generative AI reduces the time and effort criminals must expend to deceive their targets.” Criminals can create convincing images and videos of loved ones in distressing situations. These are designed to cause victims to act without thinking, as well as provide money, information and sensitive data.
Creating a secret word or phrase with your family is an important way to verify their identity. Make sure this phrase isn’t shared online.
For more advice on scam protection, staying safe online and/or spotting AI-generated content, call the Geeks on 0800 424 335. We’re here to help!
