When Anyone Can Generate the Image, What You Choose to Show Becomes Your Brand

As AI-generated hotel imagery grows, the author argues that brand integrity depends on human editorial judgment, proposing an "Arrival Test" to distinguish acceptable creative use from misleading representation.

When Anyone Can Generate the Image, What You Choose to Show Becomes Your Brand

Photo by Influence Society

The essentials

  • AI can propose endless images, but someone who knows the brand has to choose. Every image should carry your property’s signature, not a model’s generic look.

  • Before publishing, apply the Arrival Test: would a guest who booked because of this image recognize the place on arrival? Careful retouching is fine; enlarging a room or erasing a crane from a view is not.

  • Season, light, styling and people can be reimagined with AI. In the EU, a realistic AI image of a real place may now require a visible disclosure.

— 

In September 2026, a gallery overlooking the gardens of the Palais-Royal displayed a collection of rugs that took almost two years to develop1. Their patterns were born in software, then knotted by hand at around 400 knots, or seven centimeters, a day. That same season, any hotel with the right tools could turn its summer terrace into a winter scene in a single click.

Both are products of technology. Only one of them, so far, carries an unmistakable signature. The question I keep returning to is whether it has to stay that way.

The machine proposes, the maker decides

At RDAI, interior architect Johan Thuillot used Grasshopper, a parametric modeling tool, to translate the microscopic structure of butterfly wings and spiderwebs into patterns that could be rendered in Tibetan wool and Chinese silk. The software did not finish the work. His layouts went back and forth with the textile artists of Creative Matters for close to two years before artisans began knotting the first rug. “You can sense some imperfections from the hand, which adds dimension to the product,” says Anna Panosyan, one of those artists. “You can feel this is handmade.” 

Dieter Rams, now 94, has spent a lifetime on that kind of decision². Klaus Klemp, managing director of the Dieter and Ingeborg Rams Foundation, describes a designer who asks his questions before picking up a pencil. Is the object necessary? Is there already something better? How will people use it? For Klemp, “less, but better” was never a look. It is a way of thinking that comes before making.

In hospitality, the equivalent questions now come before the first prompt. Which image does this moment of the guest journey truly need? What must it tell a traveler about the place? And what should it never pretend to show?

What the data already shows 

That last question is no longer theoretical. In May 2026, the Berlin agency ABCD and the forensic verification firm ContentGuard.me analyzed 25,550 photos from 700 hotels across seven destinations3. Around 19% showed at least one sign of AI generation or AI editing, although a single signal was enough to flag a photo and the method does not separate fully generated images from enhanced ones. City hotels stood out, with 36% of photos flagged in Hamburg and 27% in Berlin, against 9% in Mallorca, a gap the authors suggest may partly reflect the landmarks and city views that fill urban galleries.

Travelers, meanwhile, cannot reliably tell the difference. In a Talker Research survey of 2,000 Americans, commissioned by the Tennessee Department of Tourist Development, only 5% picked out the real photo in all three pairs they were shown, while 74% said they would not book a trip without seeing photos first4. Seven in ten said they would place more trust in imagery from a destination that confirms it is real.

This does not mean travelers reject AI imagery outright. In a Clutch survey of 401 American consumers, 82% accepted AI imagery in at least some contexts5. Acceptance was highest for fictional scenes and concept art, and fell to 27% when AI was used to show products that already exist. Guests are not refusing imagination. They are refusing to be misled about what they will find.

The Arrival Test

Most hospitality images show a place a guest will one day discover. A room, a bathroom, a pool, a restaurant, a terrace. Treating these images as untouchable would leave little room for creativity, and treating them as raw material would break a promise. One question settles most cases. Would a guest who booked because of this image recognize it on arrival?

What must stay true

The size and layout of a room, its materials, the condition of a bathroom, the view from a balcony, the pool, the dishes that reach the table. A guest studying these details is looking for evidence. Color correction and careful retouching can help a photograph reflect what the eye would see. Beyond that, the property itself remains the reference point.

What can be reimagined, with disclosure

The season, the weather, the hour of the day, the styling, the presence of people, a pre-opening visual developed from the architect’s 3D models before a single room can be photographed. A summer photograph of a terrace reworked into a winter afternoon sets a mood without changing what the guest will find, much as a photographer has always chosen the hour and the styling of a shoot. This is where generative tools open real creative possibilities, particularly when conventional production would make an idea difficult to execute.

In the EU, however, the AI Act's transparency rules have applied since August 2, 2026. A realistic image of a real place that AI has generated or altered may need a visible disclosure, a subject we covered in detail in The EU AI Act Is Changing How Hotels Present Themselves.

How we apply it at Influence Society

AI can suggest a hundred treatments of a photograph. It cannot know on its own which one captures the quiet atmosphere of a particular courtyard, which detail the owner has spent decades preserving, or which visual exaggeration would feel alien to the brand. That is the logic behind the AI Studio we built at Influence Society, conceived as a system steered by art directors rather than a single prompt.

Our wedding film for The Six Bells shows the test in practice. The venue sent us photographs of its patio and tent.

The Six Bells' patio and tent, as provided by the venue.

From there, the AI Studio created the cast, a couple and their guests, before a single frame of the film was generated.

The cast, created in Influence Society's AI Studio before a single frame of the film was generated.

In the finished film, the tent, the curtains, the furniture, the light and the view are those of the real patio. The couple, the guests and the table setting are generated. A couple who booked the venue after watching the film would find the same space waiting for them, and each image says what was created.

Left: The Six Bells' tented patio, as photographed. Right: a frame from a wedding film created with Influence Society's AI Studio. The tent, furniture, light and view are unchanged; the couple, guests and table setting are AI-generated. 

The finished film runs 25 seconds, all of it set on The Six Bells’ real patio. 

The decision behind the image

Look again at the rugs. Software proposed the patterns, and people spent two years deciding what to keep. As generation becomes faster and cheaper, the value of that decision only grows. Hospitality brands do not need to master every new tool. They need people who understand what the brand actually stands for, and who are willing to set aside an image that does not belong to it.

When a future guest scrolls through your next campaign, they should recognize your house and your signature, not simply what a model knows how to produce.

The latest edition of SOCIETIES explores this idea across design, travel and technology.

If you would like to see how the Arrival Test applies to your own images, our AI Studio team would be glad to look at them with you.

Sources

[1] Wallpaper*, “These intricate rugs replicate the microscopic workings of Earth’s tiniest creatures” (September 2026)

[2] Sotheby’s Magazine, “Collector’s Item: Dieter Rams and the Hard Work of Simple Things” (August 2026)

[3] Breaking Travel News, ABCD Agency and ContentGuard.me analysis of 25,550 hotel photos (2026)

[4] Talker Research, “How well can Americans spot AI travel photos when booking a trip?” for the Tennessee Department of Tourist Development (June 2026)

[5] Clutch, “Think You Can Spot AI Photos? Think Again” (based on a survey of 401 US consumers, Sept. 2025)

View story source
AI in Hospitality Sales & Marketing Technology Brand Identity AI Transparency AI-Generated Imagery Hotel Photography

Personalized marketing support for extraordinary hotels - Develop your reputation and your income directly through tailor-made digital experiences. We grow compelling stories for our clients through a creative and insights-driven approach of your digital marketing strategy: brand content, social media, online advertising and mobile-first web development.

Personalized marketing support for extraordinary hotels - Develop your reputation and your income directly through tailor-made digital experiences. We grow compelling stories for our clients through a creative and insights-driven approach of your digital marketing strategy: brand content, social media, online advertising and mobile-first web development.

Comments

Comments for this content

0 comments available
Loading comments...