AIMS project
AI Incidents in Marketing
AI Incident Database // Event Marketing, AI Imagery & Misrepresentation

The chocolate factory that existed only in the prompt

AI-generated images promised families an immersive Wonka-inspired fantasy world. What ticket holders found instead was a sparsely decorated warehouse and an experience that bore little resemblance to the marketing.

Incident typeAI-generated marketing imagery, experiential marketing, misleading advertising
DateFebruary 2024
RiskConsumer deception, refunds, reputational damage, legal exposure

In February 2024, a Willy Wonka-inspired experience opened at a warehouse in Glasgow, Scotland. The event was promoted online using AI-generated images showing elaborate, candy-coloured fantasy environments and promising visitors an immersive experience filled with imaginative attractions and surprises.

Tickets cost up to £35 per person. But when families arrived, the physical experience was very different from the images used to promote it. Visitors found a sparsely decorated warehouse with limited props and backdrops, while actors hired for the event had reportedly received scripts only shortly before opening.

Parents complained and demanded refunds, the event was cancelled partway through its first day, and police attended the venue as tensions rose. Images comparing the promotional material with the actual event quickly spread online, turning the incident into an international example of the risks of using AI-generated imagery to advertise a real-world event.

What went wrong

AI image generators can produce highly polished environments without requiring the locations, sets, materials, or production budgets that would normally be needed to photograph them. This makes them useful for visualising ideas, but it also creates a significant marketing risk when imagery is presented in a way that customers may reasonably interpret as representing the actual product or experience.

Governance questions

  1. Does your marketing approval process verify that AI-generated imagery is representative of the product, service, or experience being sold?
  2. When AI-generated visuals are used as concept imagery rather than a representation of reality, how clearly is that distinction communicated to customers?
  3. Who has responsibility for checking whether the promises implied by AI-generated marketing can actually be delivered operationally?

Learning outcomes

After discussing this case, participants should be able to:

  1. Identify why disclosure of AI-generated marketing material can matter when customers may otherwise interpret the imagery as representing the real product or experience.
  2. Describe the reputational and legal exposure that can arise when marketing materially oversells a real-world experience.
  3. Assess how an organisation can verify that AI-generated creative accurately reflects the product, service, or experience being sold.

Discussion questions

  1. Would your organisation's marketing approval process catch AI-generated imagery that substantially oversells a product or experience before it goes live?
  2. Who should be accountable when AI-generated marketing materially misrepresents what a customer will actually receive?
  3. How should ticketed or experiential events manage the gap between AI-generated concept art and the experience that can realistically be delivered?