When Transparency Backfires: The Hidden Cost of AI Labels in E-Commerce
Generative AI is rapidly transforming product photography in online retail. Costs are falling, flexibility is increasing, and major players are already moving at scale. But new EU transparency requirements that came into force on 2 August 2026 are introducing a complication that no shop operator had fully anticipated: the simple act of labelling an image as "AI-generated" appears to damage product perception — and not just for AI-created visuals.
What Happened?
Two converging developments have placed product image labelling at the top of the e-commerce agenda. First, fashion retailer About You introduced a KI-based photo studio called Scayle Studios in July 2026. According to the company, production costs per item dropped from previously 70 to 80 euros per classic photo shoot to approximately one euro. About You also reported a 9.2 percent higher gross merchandise value and a 5.1 percent higher add-to-basket rate compared to conventional images. The company plans to produce all of its e-commerce imagery using AI by the end of the year.
Second, on 2 August 2026, the transparency obligations under Article 50 of the EU AI Regulation (EU AI Act) came into effect. These obligations include labelling and disclosure requirements for certain AI-generated or AI-manipulated content. At the exact moment when AI-generated product imagery is being adopted at scale, retailers are now legally required to flag it as such.
An experimental study cited in the original reporting adds a deeply inconvenient finding to this picture: the AI label reduces how customers perceive a product — and crucially, this negative effect occurs even when the photograph was actually taken by a human and is merely labelled as AI-generated by mistake or association.
The Details
The business case for AI product photography is compelling. About You's figures illustrate just how dramatic the cost reduction can be. Moving from conventional photo shoots at 70 to 80 euros per product to approximately one euro per image represents a fundamental shift in the economics of visual content production. Scaling product imagery across thousands of SKUs, seasonal campaigns, and localised storefronts becomes an entirely different proposition at this cost level.
The performance metrics reported by About You also suggest that AI-generated images are not simply a cost-cutting measure — they appear to drive commercial outcomes. A 9.2 percent increase in gross merchandise value and a 5.1 percent increase in add-to-basket rate are operationally significant figures for any e-commerce operation.
However, the new legal framework changes the calculus. Article 50 of the EU AI Act introduces labelling and disclosure obligations for AI-generated or AI-manipulated content. For retailers using generative AI in product photography, compliance now requires identifying and marking this content for consumers.
The experimental study finding is where the complexity deepens. According to the research referenced in the source article, even authentic, human-created product photographs experienced a deterioration in customer perception when the AI label was applied to them. This suggests that the label itself — independent of the actual provenance of the image — triggers a negative response in consumers.
Why This Matters for Shop Operators
For e-commerce managers and Shopware merchants, this situation creates a genuine strategic tension. On one side sits a powerful technology that demonstrably reduces costs and, in at least one major retailer's experience, improves commercial performance. On the other side sits a regulatory requirement that may actively suppress the conversion benefits that make the technology attractive in the first place.
The finding that real photos are also negatively affected by the AI label points to a broader consumer trust dynamic. The label appears to function as a signal of authenticity doubt, regardless of whether that doubt is warranted. This is not simply a compliance headache — it is a user experience and brand perception issue that will need to be actively managed.
For developers building or extending Shopware storefronts, this also raises practical questions about how and where labels are displayed, how they are integrated into product detail pages, and whether there are layout or copy approaches that can comply with the letter of the regulation while minimising the psychological impact on shoppers.
Practical Considerations
- Audit your image pipeline: Before implementing any labelling strategy, map out which images in your catalogue are AI-generated, AI-assisted, or conventionally produced. Applying labels inaccurately — or more broadly than required — may amplify the negative perception effect unnecessarily.
- Review the specific scope of Article 50: The EU AI Act's transparency obligations apply to certain categories of AI-generated or AI-manipulated content. Not every image touched by an AI tool will necessarily fall under the same requirement. Seek qualified legal guidance on what must be labelled in your specific context.
- Test label placement and phrasing: The study finding suggests that the label has a psychological effect on consumers. How, where, and in what language the disclosure is made may influence the magnitude of that effect. A/B testing different approaches — within the bounds of legal compliance — is worth prioritising.
- Don't avoid the technology: The cost and performance figures from About You are significant. Avoiding AI-generated imagery purely to sidestep labelling requirements may mean leaving substantial efficiency and revenue gains on the table.
Outlook
The EU AI Act's transparency provisions mark the beginning of a longer regulatory evolution around AI-generated content in commercial contexts. As more retailers follow About You's direction and migrate product imagery to AI-generated formats at scale, consumer familiarity with AI labels will likely grow. Whether that familiarity will normalise the label and reduce its negative perception effect — or whether it will permanently recalibrate consumer trust dynamics around product photography — remains to be seen.
What is clear is that e-commerce operators can no longer treat AI-generated imagery as a purely technical or financial decision. It is now also a compliance matter, a brand communication question, and increasingly a conversion optimisation challenge that sits at the intersection of all three. Staying ahead of that complexity requires both technical readiness and a close eye on how regulatory and consumer sentiment continues to develop.