AI Showreel consulting-grade analysis, for everyone FR
← The index
Proof C Live confirmed Failure / pullback

Willy's Chocolate Experience

AI-generated event visuals (overselling)

IndustryOtherLeverAcquisitionFamilyGenerationImplementationCustom AIStagepurchase
ferme le jour meme
Anger, refunds demanded, police called to the site
"Police were called to the venue" S1

In February 2024, the Willy's Chocolate Experience event in Glasgow sold tickets at around 35 pounds with AI-generated visuals unrelated to reality; the police were called and the event was shut down the same day.

Objective

Sell tickets for an immersive experience inspired by the Willy Wonka universe, relying on attractive visuals to convince families.

The deployment

In February 2024, an event named Willy's Chocolate Experience took place in a Glasgow warehouse. The ticketing site promised an enchanted world with gardens, characters, and shows, illustrated with AI-generated images. The visuals contained telltale signs, including made-up text such as exarserdray lollipop. On site, families who had paid around 35 pounds per ticket found a nearly empty warehouse with a few cheap props and an inflatable structure. The gap triggered anger, refund demands, and a call to the police. The event was shut down the same day. An actor who was hired described the script he received as AI-generated gibberish.

Results Proof C

ferme le jour meme
Anger, refunds demanded, police called to the site
"Police were called to the venue" S1
textes IA inventes
Sign of AI overselling in the promotional material
"exarserdray lollipop" S1

Established press and a reference institute (Poynter, Semafor, The Hollywood Reporter, New York Times cited) document the event by name, the use of AI images, and the shutdown. No published court decision at the time of verification, hence C.

How it works

Inferred typical approach

The internal detail is not public. Here is a proven approach that leads to the same result, to adapt to your stack.

reaction (colere, remboursements, police) Organisateur (marketingevenement) Generation d'imagespromotionnelles Generation du script(spectacle) Site de vente de billets Familles (acheteurs surplace)

The stack in detail

  • llm Generateur d'images IA (outil non identifie) Consumer image-generation tool used for the promotional visuals; the visible made-up text (exarserdray lollipop) reveals the generative origin. The organizer did not disclose the exact tool.
  • llm Generateur de texte IA (outil non identifie) The script provided to the actors was described as AI-generated gibberish; the specific tool is not documented.
  • plateforme Site de vente de billets de l'evenement Distribution channel for the generated visuals and ticket sales at around 35 pounds; closed after the event.

Post-mortem

Graveyard

What happened sourced

In February 2024, an organizer sold tickets at around 35 pounds for Willy's Chocolate Experience in Glasgow, illustrating the offer with AI-generated images promising gardens, characters, and shows. On the day, families found a nearly empty warehouse with cheap props. Anger, refund demands, police called, shutdown the same day. The script provided to the actors was also AI-generated and judged unusable. The affair went viral and was covered even by the New York Times.

Reason for failure sourced

The AI visuals created a promise with no connection to what could be delivered. The core problem is a gap between advertising and service. AI was used both to produce images too good to be true and a meaningless script, with no human oversight to check feasibility or coherence.

Cost sourced

Refunds demanded by families, major reputational cost for the organizer, worldwide press coverage. Precise amounts not published in consolidated form.

Warning signs inferred

Inferred: promotional visuals too perfect for a small event operation, made-up text visible in the material, and a script produced without review are clear signals. A marketing promise that no production budget could keep should have raised a flag before tickets went on sale.

Lessons in hindsight inferred

Inferred: the marketing image must stay anchored to what will actually be delivered. Using AI to paint an experience you cannot produce turns a marketing operation into deception. The simple rule: never show in promotion what the service cannot honor, AI or not.

Is the pattern still valid?

Inferred: the pattern of AI-generated marketing visuals remains valid and widespread. What failed here is not the tool but its use to oversell. The same image generator, framed on a real service and supervised, produces legitimate promotional material.

How your customers perceive this type of use

Sourced studies

Un ecart net separe les annonceurs des consommateurs : 77% des annonceurs voient l'IA positivement contre 38% des consommateurs (Yahoo/Publicis, 2024). Les mesures implicites confirment le rejet declare : en EEG, les pubs generees par IA produisent une activation memorielle plus faible que les pubs traditionnelles et sont decrites comme agacantes, ennuyeuses et confuses (NIQ, 2024). La disclosure a un effet ambivalent : elle augmente fortement la confiance quand elle est remarquee (Yahoo/Publicis), mais 27% des jeunes consommateurs disent faire moins confiance a une entreprise dont la pub est creee par IA (IAB, 2024).

77% vs 38%
Annonceurs qui percoivent l'IA positivement, contre 38% des consommateurs (2024)
72%
Consommateurs qui estiment que l'IA rend difficile de savoir quel contenu est authentique (2024)
+96%
Lift de confiance globale envers l'entreprise quand la mention IA d'une pub est remarquee (avec +47% d'attrait de la pub et +73% de credibilite de la pub) (2024)

Acceptance conditions

  • Une disclosure visible : quand la mention IA est remarquee, la confiance globale envers l'entreprise augmente de 96% (Yahoo/Publicis 2024)
  • Une qualite visuelle suffisante : les visuels IA de basse qualite augmentent l'effort cognitif et distraient du message (NIQ 2024)

Red lines

  • Le contenu IA non declare puis identifie : 72% des consommateurs disent que l'IA rend l'authenticite difficile a etablir (Yahoo/Publicis 2024) et les marques utilisant des pubs IA sont plus souvent jugees inauthentiques ou non ethiques par les consommateurs que par les dirigeants (IAB 2024)
  • Les mannequins et personnes generes par IA : 46% des consommateurs n'en veulent pas dans la publicite, l'inquietude premiere etant les standards de beaute irrealistes (Attest 2025)

Sources: Yahoo / Publicis Media (terrain Ebco) 2024 · IAB (avec Attest) 2024 · NIQ (NielsenIQ) 2024 · Attest 2025

See full acceptance: by country, by use, by generation

How to replicate

Inference, not sourced

Data prerequisites

  • descriptif reel de la prestation
  • budget et moyens de production
  • references visuelles honnetes

Org prerequisites

  • relecture humaine des visuels et du script
  • controle de conformite promesse vs livrable

Possible stack

  • generateur d'images
  • generateur de texte
  • revue editoriale et juridique

First step: Fixer d'abord ce que la prestation livrera reellement, puis ne generer que des visuels qui restent dans ces limites ; faire relire chaque visuel et chaque texte par un humain.

Sources

  1. S1 Welcome to Willy's AI catastrophe Established press poynter.org · 2024-02-29 · accessed 2026-07-11 archive pending
  2. S2 The Willy Wonka fiasco underscores the problems of AI ads Established press semafor.com · 2024-02-28 · accessed 2026-07-11 archive pending
  3. S3 Willy Wonka Event Slammed as Epic Letdown, Compared to Meth Lab Established press hollywoodreporter.com · 2024-02-27 · accessed 2026-07-11 archive pending