Skip to main content
    Step 1 of 7
    Safety & Privacy
    Beginner
    9 min read 7 stepsMay 8, 2026Verified May 2026

    How to Spot an AI-Generated Photo

    Plain-English guide to recognizing fake photos made by artificial intelligence, with simple visual checks anyone can do without special software.

    At a Glance

    Category
    Safety & Privacy
    Difficulty
    Beginner
    Read Time
    9 min read
    Steps
    7
    Topics covered
    ai
    photos
    scam
    deepfake
    social media
    seniors
    fact-check
    1

    Look Closely at the Hands and Fingers

    ~51s
    AI photo software has long struggled with human hands. Look at every hand in the photo. Count the fingers on each hand. A normal human hand has four fingers and one thumb, for five total. AI-generated hands often have six fingers, four fingers, or fingers that bend in strange directions. Look at the fingernails too: are they all the same shape, or does one nail look melted into the next? Are the knuckles in the right places? Are the wrists thicker than the rest of the arm? Hands holding objects are even trickier for AI: a fake person holding a coffee cup may have a thumb that passes through the handle, or a finger that disappears into the side of the cup. If anything about the hands feels off, the photo is probably fake.

    Quick Tip

    Even modern AI tools released in 2026 still mess up hands more than any other body part. If you only have time for one check, check the hands.

    2

    Check the Eyes, Ears, and Glasses

    ~52s
    Real human eyes have two pupils that point in the same direction and reflect light from the same source. In AI photos, the two pupils may point in slightly different directions, or one pupil may catch light from one angle while the other catches light from a different angle. Look at the ears next: do they match each other in size and shape, and do they sit at the same height on the head? If the person wears glasses, do the temples (the side arms) actually connect to the frames, or do they fade into the side of the head? Is the reflection in the lenses consistent with the rest of the room? Real human ears almost never match exactly side to side, but in an AI photo, the two ears may look identical or have wildly different shapes.

    Warning

    If the photo shows the back of the head as well as the front, check that the hair texture and ear shape match between the two views.

    3

    Read the Words in the Background

    ~46s
    AI software still struggles with written text. If the photo shows a sign, a book cover, a street name, a clock, a license plate, a t-shirt logo, or any other written words, look at the text closely. Real text uses real letters that spell real words. AI text often shows letters that look almost right but spell nonsense like "Calefe" instead of "Cafe" or "Stoip" instead of "Stop." Numbers may be backwards or have extra digits. Clocks may show impossible times like 13:67. If you see any blurry, garbled, or made-up text in a photo, it was almost certainly created by AI. This single check works on the majority of fake photos in news articles and social media posts.

    Quick Tip

    Zoom in on any text by tapping or clicking the photo to enlarge it. AI text often only becomes obvious when you look at it big.

    4

    Check Shadows, Lighting, and Reflections

    ~52s
    In a real photo, light comes from one or two sources, and every object in the scene casts a shadow in the same direction. In AI photos, shadows often point in different directions, or some objects cast a shadow while others nearby cast none. Reflections in mirrors, windows, eyes, or shiny surfaces should match what is in front of them. In AI photos, a mirror may show a reflection that does not match the room, or a person's eyeglasses may reflect a scene that is not present. The sun cannot shine from two directions at once. A real fire glows orange, and that orange glow should appear on the nearest faces and walls. If the light in the photo does not behave the way light does in the real world, the photo is probably fake.

    Quick Tip

    If a person is standing outdoors, look at the ground. A real shadow follows the shape of the body. A made-up shadow may be blurry, missing, or pointing the wrong way.

    5

    Run the Photo Through a Reverse Image Search

    ~49s
    A reverse image search lets you ask the internet: where else has this picture appeared? Open a web browser and go to images.google.com or tineye.com. Find the small camera icon in the search bar and click it. The page will ask you to either drag a photo into the box, upload a photo from your computer, or paste a web address of the photo. Once you provide the photo, Google or TinEye will search the entire internet for matches. If the photo appears in dozens of articles dated years before the supposed event, the photo is older than claimed. If the photo only appears on one shady website, it may be made up. Real news photos almost always show up on several reputable news websites at once.

    Warning

    AI photos are not always in search results because they may be brand new. If a photo appears nowhere except where you found it, that alone is suspicious.

    6

    Look for the Skin and Texture Smoothness Tell

    ~50s
    Real human skin has tiny imperfections: pores, freckles, fine lines, stray hairs, small blemishes. AI faces often look too smooth, as if they have been heavily filtered. Look at the cheeks, forehead, and chin closely. If the skin looks like polished plastic with no texture at all, the face may be made up. The same is true for fabric: real cloth has wrinkles, threads, and tiny snags, while AI fabric can look painted on. Hair is another tell: real hair has flyaway strands, while AI hair sometimes blends into the background in a soft, fuzzy way that no real haircut produces. The overall feel of an AI photo is sometimes described as too perfect, like a magazine cover from forty years ago that has been heavily airbrushed.

    Quick Tip

    Compare the photo in question to a known real photo of the same kind of scene. After a few comparisons, the AI smoothness will jump out at you.

    7

    Verify the Story Around the Photo

    ~51s
    A photo is only as trustworthy as the story attached to it. Even if a photo looks real, ask: who took it, when, and where? Real news photos always have a caption with the photographer's name, the date, and the location. If a photo shows a major news event, check whether the event was covered by major news organizations like the Associated Press at apnews.com, Reuters at reuters.com, or your local TV news station. If the event only appears in one social media post or one obscure website, the event itself may be made up. For charity appeals after disasters, donate only through trusted organizations like the American Red Cross at redcross.org, never by clicking a link in an email or a social media post.

    Warning

    If a photo is designed to make you angry or sad, slow down and check the source before sharing it. Scammers and political bad actors count on strong emotions to spread fake images faster than people can verify them.

    You Did It!

    You've finished reading: How to Spot an AI-Generated Photo

    How well did this guide stick with you?

    Need more help? Book a TekSure tech

    Photos used to be reliable proof that something happened. If you saw a photo of a person standing in front of a famous building, you could trust they were really there. In 2026, that is no longer true. Free websites can now create lifelike photos of any person, place, or event in seconds, based only on a written description. These are called AI-generated photos, or synthetic images. They show up in news articles, on Facebook, in scam emails asking for charity donations, and in fake stories meant to make people angry or afraid. Federal Bureau of Investigation reports show AI image fraud cost Americans over $2 billion in 2025, with the largest growth coming from romance scams and charity fraud aimed at seniors.

    The most common ways seniors run into AI-generated photos are these: a fake romance interest sends pictures that look like a model but turn out to be made up; a charity scam after a hurricane shows children in rubble who never existed; a political post shares an outrageous photo of a politician that turns out to be fabricated; an obituary scam uses an AI photo to make a fake death announcement look real, then asks for funeral donations. The technology has improved so fast that even careful viewers can be fooled. But fake photos still have a handful of giveaway flaws that you can learn to spot in under a minute.

    The good news is that you do not need special software or technical training to catch most AI photos. The same eyes that helped you sort real things from fake your whole life still work, you only need to know what to look for. AI photo software is very good at making big shapes look right, like faces, bodies, and backgrounds. It is much worse at small details, especially hands, ears, jewelry, eyes, and written text. Looking at those small details is the fastest way to tell a real photo from a fake one. This guide shows the seven checks that work most of the time, in the order you should do them.

    Before you begin, find a photo you want to check. It might be one from a news article, an email, a Facebook post, or a dating app profile. Have it open on your screen large enough that you can see the small details. If the photo is too small, click or tap it to make it bigger. Have a notepad handy. None of these checks require any download or any special tool, only your eyes and a careful look. The whole process for one photo takes about two minutes once you have practiced it a few times.

    Rate this guide

    How helpful was this guide?

    ai
    photos
    scam
    deepfake
    social media
    seniors
    fact-check

    Official Resources

    Sources used to create and verify this guide. View all sources →

    Still stuck? Let a pro handle it.

    A real person can walk you through this over the phone, anywhere in the US. If we can't fix it, you don't pay.