AI-generated visuals are raising tough questions for businesses – and for consumers. People say they don’t trust AI images. They say they want authenticity. They say it matters whether an image was created by a person or a machine.
But actual behavior tells a more complicated story.
Recent poll results may show a disconnect between consumer attitudes and consumer behavior. These were informal LinkedIn polls, not scientific studies – but each showed a clear directional preference. Together, they offer useful signals for how to approach the use of AI-enhanced images in product marketing.
While public discourse suggests growing discomfort with AI-enhanced imagery, actual shopping habits reveal a more passive engagement with the issue. This gap isn’t just interesting – it’s instructive. It offers clear guidance for marketers and content creators navigating this new visual frontier.
Consumers say they want authenticity – but rarely verify it
In one poll, 79% of respondents said they didn’t try to discern whether a product image was AI-enhanced during their most recent online purchase. That’s despite the fact that, in a separate survey, a large majority also say they want to know when an image has been generated by AI.
This points to a divide between theoretical objections and practical behavior.
In many cases, consumers may be operating on autopilot – relying on brand familiarity and visual plausibility rather than investigating whether what they see has been digitally created. AI-generated images, when done well, don’t necessarily raise red flags. They blend into the visual language of e-commerce, lifestyle content, and advertising without disrupting the experience.
For brands, this doesn’t mean transparency is irrelevant. Quite the opposite – labeling or disclosing AI use in a thoughtful way can serve as a trust-building move, even if consumers aren’t demanding it in the moment. The opportunity lies in preemptive clarity, not reactive justification.
The takeaway: if an image feels helpful, natural, and context-appropriate, most consumers won’t pause to dissect how it was made. But if that same image causes confusion or disappointment later, that’s when scrutiny shows up.
Expectation mismatch still triggers returns – even if AI isn’t the issue
Another poll revealed that 67% of people would return a product if it looked different in person than it did in its images. This underscores a key truth: what matters most to consumers isn’t how an image is made – it’s whether it accurately reflects the product.
That intention should be taken seriously. A significant portion of shoppers do follow through when visuals create false expectations, especially for appearance-driven purchases or when return processes are simple. At the same time, this is where the attitude-behavior gap emerges. Not everyone who says they would return a mismatched item actually does – some will accept minor deviations or avoid the hassle.
The message for brands is clear: consumers notice when images overpromise, and many will act. AI-enhanced imagery can still play a valuable role – but only when it supports, rather than distorts, the reality of the product. A hybrid approach that improves clarity and consistency without crossing the line into misrepresentation is the most sustainable strategy.
Mind the gap, but don’t fear it
We should keep these two possible contradictions in mind. One is that consumers generally don’t inspect for AI in product images, even though the vast majority want to know when an image is made with AI. The other is that even though most consumers say they’d return a different-looking product, their actions may be different.
This attitude-behavior gap isn’t a loophole – it’s a reality. For teams using AI in visual marketing, the goal isn’t to exploit the gap, but to understand it and design accordingly.
Here’s what that looks like in practice:
- Label AI visuals when it matters – Not every image needs a disclosure, but critical product visuals should align with customer expectations. Use AI to improve clarity, not to mask reality.
- Enhance without exaggerating – A cleaner background or adjusted lighting can make images more appealing without misleading. Avoid enhancements that alter core attributes of what the customer will receive like color, size, or finish.
- Audit your product imagery – Regularly review how your visuals align with in-person customer experiences. If returns cluster around certain visuals, the imagery may be the culprit.
- Respect visual accuracy for key decision points – For high-consideration products, ensure the customer sees what they’ll get. That’s where trust is built – and where disappointment can do the most damage.
AI-powered visuals aren’t inherently problematic. But their effectiveness depends on how and where they’re used.
AI can still support better storytelling – and better outcomes
The debate over AI in marketing often centers on what’s real. But what’s real in storytelling has never been purely about raw documentation. It’s about resonance. It’s about setting expectations – and meeting them.
Used thoughtfully, AI can help teams create more consistent, compelling, and scalable visual experiences. And when those experiences reflect the product accurately, the result is not just higher performance – it’s longer-term trust.
To explore how AI is transforming visual media and the future of content creation, watch the on-demand webinar “How AI is Shaping Storytelling and Visual Media”. In this candid conversation, journalist Connie Guglielmo and Imgix CEO Chris Zacharias explore how creators and businesses can embrace AI in ethical, responsible, and surprisingly human ways.
