When AI Gets Your Brand Wrong

Ahead of a new initiative from Spotlight Communications and Make Lemonade focused on AI visibility and luxury travel discovery, Lucy Clifton and Maria Sze discuss why editorial authority and narrative control are becoming critical battlegrounds. After all, what’s the point of a hotel spending millions repositioning its brand, only for AI systems to continue describing it through outdated articles, abandoned listings and old narratives buried online? Lucy Clifton reports.

Lucy Clifton – Let’s start with the scale of the issue. What is AI actually getting wrong about luxury brands?

Maria Sze – More than most people realise. A lot of brands assume AI is pulling from their current website or latest press coverage. It isn’t. These systems build understanding from layers of information gathered over years, sometimes decades, and luxury brands evolve constantly. Ownership changes. Positioning changes. Restaurants close. Entire guest experiences are redesigned. But the older version often stays online indefinitely.

Lucy – So the internet never really forgets.

Maria – AI certainly doesn’t. And honestly, for a lot of luxury travel brands, it’s not going terribly well. We’ve seen hotels being described through facilities they removed years ago. Resorts positioned around concepts they abandoned completely. A hotel can spend millions repositioning itself while ChatGPT is still describing the version that existed three owners ago.

Lucy – Which is slightly terrifying when you think how many people are now planning trips through AI.

Maria – Completely. And most brands have absolutely no visibility into how they’re currently being interpreted by these systems. They only realise something’s wrong when a guest repeats back an odd description or a travel advisor asks a strange question.

A hotel can spend millions repositioning itself while ChatGPT is still describing the version that existed three owners ago.

When the internet doesn’t forget

Lucy – What’s interesting is that the problem sits across two completely different worlds.

Maria – Yes. One side is technical: tracing where inaccurate information actually lives online and understanding why AI systems keep retrieving it.

Lucy – And the other is editorial and reputational. Correcting the narrative through trusted publications, journalists and authoritative media channels. Neither side really solves the problem on its own.

Maria – Exactly. In the end, we kept coming back to the same thing: find the source, correct the record, then build the narrative you actually want AI to find next.

Lucy – Because if you simply delete inaccurate information without replacing it with something stronger, the vacuum gets filled again.

Maria – Usually by another version of the wrong story.

Lucy – Which is how this initiative really came about. This was all new to me 12 months ago, but over the past year we’ve spent a huge amount of time analysing how AI systems interpret luxury travel brands and destinations. What kept surfacing was the same underlying problem again and again: brands were investing heavily in repositioning themselves while AI systems continued describing them through outdated articles, archived reviews, obsolete listings and fragmented information buried across the internet.

Maria – And luxury is particularly vulnerable to that because perception matters so much. These brands are built on positioning, nuance and reputation. If AI gets the narrative wrong, it changes how a brand enters the consideration set before a guest even reaches the website.

Lucy – Which is really the important shift here. The guest journey is changing.

Maria – Completely. People aren’t really searching in the old way anymore. They’re asking questions. Which hotel feels discreet? Which resort has the best wellness programme? Which property is actually worth the money? AI systems are increasingly shaping perception before the brand interaction even begins.

AI systems are increasingly shaping perception before the brand interaction even begins.

Finding the source

Lucy – So what does “finding the source” actually involve in practice?

Maria – We audit what major AI systems are currently saying about a brand across ChatGPT, Perplexity, Google AI Overview and others. Then we trace every inaccurate, outdated or distorted narrative back to where it originates. Sometimes it’s a forgotten press release. Sometimes it’s Wikipedia. Sometimes it’s an old travel blog, a dead listing site or a structural issue on the brand’s own website.

Lucy – And often brands already suspect something is off. They’ve heard strange things from guests or noticed AI descriptions that feel slightly skewed. What they don’t know is where the problem starts.

Maria – Exactly. The audit effectively creates a narrative map: where the misinformation lives, why AI systems trust it and how it keeps resurfacing.

Lucy – And then each layer requires a different response.

Maria – Yes. There’s the brand’s own ecosystem: website copy, structured data, outdated landing pages, technical architecture. Then there’s third-party editorial coverage that no longer reflects reality. Then there are open-source knowledge systems like Wikipedia, listing platforms and aggregators.

Lucy – Which is really where Spotlight and Make Lemonade started complementing each other quite naturally.

Maria – Because you need both sides. You need the technical analysis, understanding why AI keeps retrieving the wrong information, but you also need the editorial relationships capable of replacing that narrative with something more authoritative.

Storytelling, not just SEO

Lucy – Ultimately this is still a storytelling problem as much as a technical one.

Maria – Completely. AI systems need stronger replacement signals. You can’t just remove the wrong narrative and hope the problem disappears. You have to ensure there’s clearer, more current and more authoritative information for the systems to retrieve instead.

Lucy – Which changes how you think about PR and media placement altogether. It’s no longer only about reach or impressions. Increasingly it’s also about whether coverage is structured, authoritative and visible enough to shape AI-mediated discovery over time.

Maria – And not just through obvious prestige titles. One of the fascinating things we found through the research is that AI systems repeatedly draw from a much broader ecosystem of specialist and mid-tier sources during luxury travel queries.

Lucy – Which means outreach becomes much more intelligence-led.

Maria – Exactly. We already know which publications AI systems trust for different categories of luxury travel recommendation. That changes the strategic value of media entirely.

Lucy – And this matters for almost any luxury brand going through change. Hotels repositioning after refurbishment. Resorts evolving their wellness or dining offering. Groups going through ownership changes. Destinations entering new markets. Brands whose current positioning simply isn’t reflected accurately online anymore.

Maria – Honestly, most luxury brands should probably be asking a very simple question: what does AI currently think we are?

Lucy – Because the answer may not be the version they think the world is seeing.

Maria – Often it isn’t.

Lucy – And increasingly, that disconnect is becoming commercial, not just reputational.

Maria – Yes. Because guests aren’t only discovering brands through search anymore. They’re discovering them through interpretation.

Where to start

Lucy – So where does a brand start?

Maria – Usually with an AI visibility and citation audit. Before you can shape the story, you need to understand which version of it the machines already believe.

Lucy – Which means the real danger isn’t invisibility at all.

Maria – It’s being understood through a version of yourself that no longer exists.

The real danger isn’t invisibility. It’s being understood through a version of yourself that no longer exists.