Blazej Kunke | March 2026
Just because you could spot an AI-generated image in 2024, do not assume you still can in 2026. The tells are gone. The blur, the wrong hands, the uncanny skin? All fixed.
What Changed
A year ago, AI-generated images had reliable fingerprints. Extra fingers. Plastic skin. Backgrounds that dissolved into meaningless blur. Text that looked like it had been through a washing machine. Trained eyes could spot them. There was a workable, if imperfect, system of visual trust.
That system no longer works. The new generation of image models, including Google’s Nano Banana 2, produces photorealistic results that are functionally indistinguishable from real photographs. The architectural leap is not incremental. These models reason about composition, lighting, anatomy, and context before rendering a single detail.
To test this firsthand, I used Nano Banana 2 to reimagine myself as someone of Japanese origin. The result is below. I would challenge anyone to identify it as AI-generated without prior knowledge that it is.
Why This Matters for Business
For most business owners, the immediate reaction is to think about marketing. And yes, the implications for visual content production are significant. What once required a photographer, a model, a studio, and post-production can now be done in seconds. That is both an opportunity and a disruption, depending on which side of the equation you sit on.
But the more important implication is about trust. Visual evidence has long been one of the foundations of credibility. A photo of a product, a team, an event, a location. These carry weight precisely because they feel real. That weight is now in question.
Consider what this means practically. A competitor’s glowing customer testimonial with a photo. A supplier’s factory images. A partner’s team page. A candidate’s LinkedIn profile photo. None of these can be taken at face value in the same way they once could. This is not a distant problem for media organisations. It is a present-day operational reality for any business that relies on digital communication.
The Opportunity Side
It would be wrong to frame this purely as a threat. The same capability that complicates visual verification also unlocks creative and commercial possibilities that were previously inaccessible to small and mid-sized businesses.
Localising marketing visuals for different markets without reshooting. Prototyping product imagery before manufacturing. Generating consistent brand visuals at scale without agency budgets. Testing different visual identities in hours rather than weeks. These are real advantages, and early adopters will move faster than those who wait for the technology to feel comfortable.
What to Do Now
There is no single answer, but a few priorities are worth keeping in mind.
Raise your verification standards. Do not assume visual content is authentic. This applies to third-party content you receive as much as content you evaluate in the market.
Invest in authentic documentation. Real photos from real events, real people, real workplaces carry increasing value precisely because AI-generated alternatives are now so accessible. Authenticity is becoming a differentiator.
Understand the tools before your competitors do. You do not need to become a Machine Learning engineer. But understanding what these models can and cannot do gives you a clearer picture of where the risks and opportunities actually lie.
The Ground Has Shifted
The gap between a real photograph and a generated one is now effectively zero. That is a remarkable technical achievement. It is also a genuine shift in how we should think about visual information, not just as consumers, but as business owners and decision makers.
The question is no longer whether AI can produce convincing images. It can. The question is what you are going to do with that knowledge.