Don't fall for artificial intelligence deepfakes: Here's how to spot them (2024)

Don't fall for artificial intelligence deepfakes: Here's how to spot them (1)

Tom Hanks isn’t trying to sell you a dental plan. YouTuber MrBeast won’t give you an iPhone 15. CBS News host Gayle King isn’t recommending a weight-loss product.

I knew this stage of artificial intelligence tomfoolery was coming, but it’s still surprising how fast it’s happening. Let’s take a closer look at how free and cheap tools are fueling fraud – and the signs to watch for.

I write about the latest tech gear, security alerts and digital life hacks in my daily emails.Join 600K+ who get the Current each day. It’s free!

Beyonce, Beytwice

Over the past several months, a whole crop of fake AI celebrities has sprung up to trick people with false endorsem*nts. They look and talk exactly like trusted personalities, and they’re usually shilling brands you’ve never heard of. Worse, they’re almost impossible to stop.

Certain AI – including Stable Diffusion apps and some Nvidia tech – can easily mimic “big names” in the celebrity world. Unscrupulous brands, or outright scammers, have no problem taking advantage of this readily available new tech.

The fight against fake photos:How Adobe is embedding tech to help surface authenticity

Elon Musk, Barack Obama, Tom Brady, Beyonce – the list keeps going, and it’s not just about how celebrities look. AI also excels at mimicking famous voices, which is how fans mistook songs by AI mixer Ghostwriter as new releases from Drake and The Weeknd. Feeling paranoid?

Keep your guard up

Here’s what social media sites and even Google don’t want you to know: They do a lousy job vetting advertisers. They’re focused on the cash, meaning scammers end up with the real results.

If you want advice about a product or service, look at reviews or find an expert who’s well-versed and can prove it. Another smart step: Google the product and actor in the ad, along with the word “review.” If someone’s getting paid to endorse a product, there’s not just one random ad floating around social media.

Tells that it’s AI-made

It’s not just ads. AI is everywhere, and the tools are simple to access and widely accessible. Take a look at what Michael Jackson would look like today, Elon Musk as a toddler, or if selfies existed in history's biggest moments.

I used one of those viral AI headshot apps. My favorite shot is the one with my fingers going into my cheeks. Now is the time to brush up on the red flags a graphic, photograph, piece of art, or image was made using AI.

Perfect symmetry: AI takes symmetry to a whole new level. In some cases, everything looks unnaturally balanced – to the point that you could fold an image in half, and it would be perfectly mirrored on both sides. Landscape photos and close-up shots are prime candidates.

Distortion: There’s usually a little wonkiness or distortion hiding in an AI image. Look at the faces in a crowd; chances are you’ll find someone who looks like they went for a swim in an oil painting.

Unusual proportions: Objects might seem too big or small, like cars as tall as homes or doorways too tiny for anyone to squeeze through.

Shadows and light: Real-world photos play nice with the laws of physics, but AI? Not so much. Shadows tend to go in all directions.

Unnatural textures: Images might appear too smooth, with no natural textures or color variations. On the flip side, it can have the opposite effect; hyper-realistic portraits with zoom-worthy wrinkles, pores and blemishes are also a thing.

Watermarks: Most free AI image generators stick watermarks onto their creations. You’ll usually find them in the bottom right corner in the form of a logo or text pointing to the site or app.

Weirdness: While AI continues to improve, there’s still a trace of the bizarre. Look for oddities like strangely shaped features, too many fingers, or accessories that don’t match. It’s always the small things!

Tools you can use

If you’re still feeling iffy about an image’s origin, try a free AI detection tool. They’re not perfect, but they can help.

AI or Not: Simply drag and drop, upload, or paste an image URL into the search bar, and AI or Not will tell you if it's legit.

Maybe's AI Art Detector: Pretty much the same as AI or Not, but it will give you the AI verdict in percentages (like “20% human and 80% artificial”).

Don’t forget about Google’s reverse image search. If Google returns with the same image from credible sources, chances are it's real. But if it points you to an AI site, you might want to sound the alarms. Here’s how to check on iPhone and Android.

Keep your tech-know going

My popular podcast is called “Kim Komando Today.” It’s a solid 30 minutes of tech news, tips, and callers with tech questions like you from all over the country. Search for it wherever you get your podcasts. For your convenience, hit the link below for a recent episode.

PODCAST PICK:America's UFO hotspot, voice cloning scams & incognito mode tricks

Plus, not a fan of smiling for photos? Google's new app can turn that frown upside down. Caller Matt thinks he subscribed to The Weather Channel but is uneasy about it needing his Social Security number. Also, the lowdown on state e-bike rebates, Klarna's shopping AI and how to spot a hacked webcam.

Check out my podcast “Kim Komando Today” on Apple, Google Podcasts, Spotify, or your favorite podcast player.

Listen to the podcast here or wherever you get your podcasts. Just search for my last name, “Komando.”

Learn about all the latest technology on theKim Komando Show, the nation's largest weekend radio talk show. Kim takes calls and dispenses advice on today's digital lifestyle, from smartphones and tablets to online privacy and data hacks. For her daily tips, free newsletters and more, visit her website.

Don't fall for artificial intelligence deepfakes: Here's how to spot them (2024)

FAQs

Don't fall for artificial intelligence deepfakes: Here's how to spot them? ›

Things like unnatural blinking, skin texture or strange lighting can all be tell tale signs that something isn't right. Audio deepfakes might also sound a little robotic, or have unnatural fluctuations in pitch. However, the best way to spot deepfakes is through being informed about how the technology is evolving.

How to spot AI deepfakes? ›

Facial and body movement

For images and video files, deepfakes can still often be identified by closely examining participants' facial expressions and body movements. In many cases, there are inconsistencies within a person's human likeness that AI cannot overcome.

How to spot fake AI images? ›

Experts advise looking closely at the edges of the face. Does the facial skin tone match the rest of the head or the body? Are the edges of the face sharp or blurry? If you suspect video of a person speaking has been doctored, look at their mouth.

How to protect yourself from AI deepfakes? ›

Use Watermarks on Photos:

Visual media can be easily manipulated with AI apps. You should consider adding a watermark so your pictures can serve as a deterrent against potential deepfake manipulation.

How to counter AI deepfakes? ›

To combat such abuses, technologies can be used to detect deepfakes or enable authentication of genuine media. Detection technologies aim to identify fake media without needing to compare it to the original, unaltered media. These technologies typically use a form of AI known as machine learning.

How to spot AI text? ›

Abrupt shifts in tone, style, or topic can point to AI that is struggling to maintain coherent ideas. Whereas humans often vary structure to create a better flow, repeated phrases or sentence structures can point to AI relying more on memorized patterns.

What is a real life example of a deepfake? ›

A well-known example of political deepfake misuse was the altered video of Nancy Pelosi. It involved slowing down a real-life video clip of her, making her seem impaired. There are also instances of audio deepfakes, like the fraudulent use of a CEO's voice, in major corporate heists.

Can deepfakes be illegal? ›

Laws are in place to take legal action against deepfakes that ruin reputation, bully, or violate privacy. If deepfakes spread false and harmful information causing defamation, libel laws can be invoked.

How can deepfakes be stopped? ›

1. Technology. Multiple technology-based detection systems already exist today. Using machine learning, neural networks and forensic analysis, these systems can analyze digital content for inconsistencies typically associated with deepfakes.

What to do if you are a victim of deepfakes? ›

Report deepfake content: If you come across deepfake content that involves you or someone you know, report it to the platform hosting the content. This can help in having it removed or investigated, limiting its potential reach. You should also report it to federal law enforcement.

How do I block my AI? ›

Find 'My AI' under Recent Conversations, then use the toggle to disable My AI.

How do I hide AI detection? ›

Techniques for How to Avoid AI Detection
  1. Using Unicode Characters. ...
  2. Adding Punctuation and Symbols. ...
  3. Using hom*oglyphs. ...
  4. Using Synonyms and Antonyms. ...
  5. Rearranging Sentence Syntax. ...
  6. Changing Word Forms. ...
  7. Visual Camouflage Techniques. ...
  8. Audio Encoding Methods.
Apr 23, 2024

How to avoid deep fakes? ›

Consumers must be cautious online, verify sources, and be aware of the potential harm caused by deepfakes. One way to identify deepfakes is with deepfake protection tools that use a variety of sophisticated techniques to detect and mitigate the risk of media content that's been manipulated.

How can you tell fake faces generated by AI? ›

AI-generated images are getting better and better, but there are still some telltale signs you can check.
  1. Hands and limbs. Most people have five fingers on each hand, two arms and two legs. ...
  2. Words. ...
  3. Hair. ...
  4. Symmetry. ...
  5. Textures. ...
  6. Geometry. ...
  7. Consistency. ...
  8. Don't get hung up on AI.
Apr 8, 2024

How can you tell if someone is AI-generated? ›

How to spot artificial intelligence-generated content
  1. Find a high-resolution image and zoom in.
  2. Look for asymmetry and other physical inconsistencies.
  3. Strange textures or a glossy effect.
  4. Look out for strange shadows, blurs, or light flickers.
  5. Unnatural body language.
  6. Take a closer listen.
Dec 14, 2023

How to spot generative AI? ›

AI detectors algorithmically apply much the same logic when identifying patterns and probabilities relating to AI generation.
  1. Consistency versus randomness. ...
  2. Repetitive words and phrases. ...
  3. Incoherence. ...
  4. Incorrect grammar and punctuation. ...
  5. Inaccuracies. ...
  6. Lack of sourcing. ...
  7. Devoid of personality. ...
  8. Doesn't pass the smell test.
Aug 2, 2023

How do you identify AI-generated art? ›

AI mimics the final product of an artwork, but does not understand its construction. Actual artworks will look layered in a way. An AI cannot put different layers of color onto a piece. This results in a piece that seems somewhat blended— there are no fully crisp lines.

References

Top Articles
Latest Posts
Article information

Author: Ray Christiansen

Last Updated:

Views: 5865

Rating: 4.9 / 5 (49 voted)

Reviews: 80% of readers found this page helpful

Author information

Name: Ray Christiansen

Birthday: 1998-05-04

Address: Apt. 814 34339 Sauer Islands, Hirtheville, GA 02446-8771

Phone: +337636892828

Job: Lead Hospitality Designer

Hobby: Urban exploration, Tai chi, Lockpicking, Fashion, Gunsmithing, Pottery, Geocaching

Introduction: My name is Ray Christiansen, I am a fair, good, cute, gentle, vast, glamorous, excited person who loves writing and wants to share my knowledge and understanding with you.