In today’s digital world, we have to think twice about everything we see. This is because the number of fake images, videos, texts, and other forms of content is growing every day.
What’s behind all of this? Artificial intelligence software.
As these tools continue to evolve, navigating the internet and social media becomes increasingly difficult. Today, AI can generate content so advanced that it is almost impossible to tell whether something is real or not. And this is only going to get worse, especially when it comes to photos and videos.
In August 2025, Google launched Nano Banana, an image generator powered by Gemini. Nano Banana has the ability to create incredibly accurate and believable images, even more so than other similar platforms.
So what makes Nano Banana different from tools like Sora, Grok, or other image generators?
Mainly two things.
On the one hand, it stands out for its ability to read a text description and generate an image that is extremely faithful to what is requested. This is something many other tools have not yet managed to fully master.
On the other hand, it also excels at editing existing images, adding elements or changes that are almost impossible to detect as AI-generated, even for professionals.
Although in recent months Gemini has begun to limit certain uses of Nano Banana (for example, it refuses to generate images featuring public figures to prevent deepfakes), the images it produces are still very difficult to distinguish from real content.
We decided to test it ourselves to see just how accurately it could describe and recreate a scene.
With everything currently happening in the United States regarding ICE, we asked Nano Banana to create a street protest scene. We requested a large crowd, police presence, and specified Minnesota as the location, where these events are unfolding.
The initial image was already believable. As we continued to request changes and add more details, the scene became increasingly realistic.
The final result (the image shown) is almost impossible to identify as fake, especially considering that this type of content is usually viewed quickly and without much context.
The fact that we were able to create a completely false scene so easily, and one that many people could believe is real, is truly alarming. Anyone could do something similar at any time and ask AI to generate almost any situation.
There is also an important point to consider: we are not experts in creating AI-generated images.
Now imagine what someone with experience and malicious intent could do.
Today, with technology that anyone can access in just a few seconds, it is becoming increasingly difficult to know whether the images, videos, or stories we see online are real or AI-generated. That is why it is essential to be careful not only when consuming content, but also when sharing it.
Confusion begins when we stop questioning what we see. The issue we keep coming back to is the same: it is no longer easy to know what to trust.
In an environment where fake content can look completely real, relying on intuition alone is no longer enough. We need tools that help verify the authenticity of digital content.
Discover Vericta, a solution designed to detect whether images and videos have been generated or manipulated using artificial intelligence, helping restore trust in online content.
