I would like to give you tips how to recognize AI generated images. They are getting more realistic, but they still tend to “cheat” on details, physics and text, especially in emotionally charged viral posts. If you know where to look, you can spot many of them in a few seconds and you can avoid sharing fake nostalgia photos, miracle stories or political information that seems true.

Note: all of the images below are AI generated.

1. Start with a quick vibe check

Before you zoom into pixels, ask a simple question: does the overall scene make sense in the real world?

• Is the story almost too perfect, dramatic or heartwarming to be true?
• Does the image hit your emotions first and only later your critical thinking?

AI images shared on social media often target nostalgia, religion or empathy, because these themes lower our guard and trigger cognitive biases like confirmation bias and familiarity.

2. Count fingers, teeth and limbs

The second way to recognize AI generated images is looking at anatomy. Anatomy is still a weak spot for many models, especially in busy scenes.

Look for:

• extra or missing fingers, fused fingers, or fingers bending in physically impossible ways
• strange hands: too big, too small, wrong angle, fingers going through objects
• teeth that form a continuous “white wall” instead of individual teeth
• too regular a row of teeth / the people in the picture have almost the same row of teeth

These errors are more common in group photos, dynamic poses or images with many overlapping limbs.

3. Inspect faces and symmetry

AI can generate beautiful faces, but they often become slightly unnatural when you look closer.

Check for:

• Eyes looking in slightly different directions, different eye sizes, off‑center pupils
• Ears that are distorted, missing, blurred into hair, or have no clear structure
• Hairlines that look painted on or impossible curls and braids

Another tell: faces that look too symmetrical and flawless, almost like a beauty filter turned to maximum.

4. Look for background and object glitches

AI tends to “give up” on background details, especially when the foreground is complex.

Look for:

• objects that float, merge, or intersect in impossible ways (a mug melting into the table, a chair leg disappearing, faces are distorted etc.)
• reflections in windows, mirrors or water that don’t match reality
• background people with melted faces, broken glasses or missing limbs

These anomalies often hide at the edges of the frame. Zooming or mentally walking through the scene helps.

5. Study textures and materials

Many models still struggle with realistic surfaces, especially when patterns or fine details are involved.

Suspicious signs:

• skin that is too smooth, plastic‑like or uniformly perfect, with no pores or subtle imperfections
• clothing with chaotic folds or fabric that doesn’t follow gravity, or patterns that bend in impossible ways
• repeating textures where there should be variation (tiles, grass, bricks, crowd clothing)

6. Check text, logos and signatures

Even that image models are more and more precise, text is a classic weak point: AI often produces spelling errors, warped letters or meaningless symbols.

Watch out for:

• street signs, shop names, T‑shirt prints, posters and book covers with garbled or unreadable text
• signatures on the corner of an artwork that look smeared, inconsistent or unreadable
• logos that are almost right, but with wrong font, missing elements, mirrored details

Even with newer models, small text in the background is still a good place to look for glitches.

7. Test the physics: light, shadows, gravity

Reality has rules. AI sometimes ignores them.

Look for:

• shadows going in different directions or not matching the visible light sources
• liquids that look solid or frozen in an unnatural way (coffee that doesn’t move, water that behaves like gel)
• objects that seem to float or rest in unstable positions without any support
• the creases in the clothes are strange

If the lighting feels movie‑perfect everywhere:

• every face perfectly lit,
• no harsh contrast,
• no real imperfections

that can also be a subtle sign.

8. Look at crowds, pets and complex scenes

AI is much better at single portraits than at big, messy situations.

Potential tells in crowds:

• background people with distorted faces, extra limbs or missing body parts
• animals with strange anatomy: too many legs, missing tails, weird eyes
• repeated clones of the same person or dog in different parts of the image

Group scenes with many interactions (hugs, handshakes, sports) are still a frequent source of glitches.

9. Watch for over‑polished, style‑averaged art

In illustrations and digital paintings, AI often produces beautiful, but generic look.

Typical signs:

• many different stylistic influences mashed together without a strong personal style
• lighting and rendering that are extremely glossy and perfect, with no sign of brushwork or consistent technique
• backgrounds that feel less carefully designed than the main character

If an artwork looks like ArtStation, Pinterest and Instagram blended into one, but you can’t feel an individual artist’s hand, it might be AI.

You can notice on AI digital paintings also similar signs which I mentioned above (like distorted text, distorted faces).

10. Combine visual clues with context

Finally, don’t stop at the pixels. Context still matters.

Ask:

• Where is the image posted? Does the account frequently share AI art or sensational content?
• Does the caption mention an AI tool, or use hashtags like #aiart, #midjourney, #aigenerated?
• Is there any source, photographer name, place or date – or is everything vague?

If you are still unsure, you can look at metadata or use reverse image search, but even without tools, these simple questions often reveal that something is off.

Summary

We know that genAI models are constantly evolving. Sometimes it’s almost impossible now to make a difference between real and fake art or photo.

I hope this guide helps you to notice the differences. You can train your eye, you can create AI images yourself and see the difference between photo and generated visual content.

You can also search for “fake or photo quiz” or “AI or Photo” and you can take quizzes – this is fun and also a good training opportunity.

The post How to recognize AI generated images without special tools? appeared first on AI Art Blog.

What are the most important changes?

The new software includes the 3 previous ones in one app.

You know that up to and including Affinity v2, these 3 software can be used separately:
Affinity Photo – Photo editing and digital painting
Affinity Designer – Vector graphics
Affinity Publisher – Publication editing

The new Affinity includes these in one. Switching between the 3 (or even more) so-called Studios is possible in the upper left corner.

The new Affinity also includes Canva AI, but its use is completely optional.

In addition, the app has Slice, Retouch and Colour Grading studios.
You can also create your own Studio, which can be completely customized.

You can find the most frequently asked questions on the official website, here is an excerpt:

Can I still use the v1 and v2 software?
Yes, support will continue and the activation server will remain available, but new licenses for v1 or v2 can no longer be purchased.

Am I able to open and edit the files created with Affinity v1 and v2 in the new app?
Yes, the new desktop version can open them.

Can I re-download my previously purchased Affinity products?
Yes, you can download them by signing-in to your account.

New features in nutshell

• pigment blend – a different type of transition option for gradients
• mesh fill type – a mesh can be used to fill shapes with a gradient
• an adjustment brush has been added to Pixel studio
• live filter brush – live filters can be painted directly on the image with a brush
• new filters: glitch – aberration, distortion, pixelate and similar effects can be added to the image in a non-destructive way
• A big new feature in Layout studio is that you can vectorize pixel images (this is called tracing, which has existed in Corel Draw for a long time, for example)
• in general, it is easier to find the panels we need (panel search function)
• quick menu: right click on the pasteboard to show it

A brief overview of AI Studio:
general AI functions are available, such as:

• generative expand
• generative fill
• generative edit
• generate image
• select subject
• remove background
• portrait blur and lighting

AI studio is available with Canva Premium Plan.

No internet connection is required to use the new Affinity, only when logging in for the first time and downloading the app.

Personally, I still don’t fully understand how they made this great software free.
Which partly explains this decision, that it seems that the AI ​​studio will be a paid service.

In the first look video, further developments were also briefly introduced:

• live blend tool
• creating custom scripts (programming)

Conclusion

Overall, these new features are interesting and I definitely recommend trying the new Affinity.

The new app can be downloaded from https://www.affinity.studio/ for Windows and macOS users. We should wait a little for the iPad app.

For your information: when you first launch the app you should decide whether you want to share usage data (turned off by default):

I created the cover image with the new Affinity desktop app. (I used a custom font with glitch filter and one layer with motion blur, other layer with gaussian blur effect)

The post New Affinity is available for free appeared first on AI Art Blog.

In a previous development of Luminar Neo the GenErase feature was introduced. GenErase can easily remove an object from an image and fills that area with relevant visual content. This also saves a lot of time when post-processing photos.

The post Skylum announced GenSwap in Luminar Neo 1.16.0 appeared first on AI Art Blog.