Meta
AI chatbots styled after celebrities
Meta launched celebrity-styled AI chatbots such as Kendall Jenner and Snoop Dogg in September 2023, then pulled them in July 2024 for lack of audience, despite multi-million-dollar deals.
Objective
Increase engagement on Meta's apps with chatbots embodying celebrity-inspired personas (voice, style), to make people want to chat with the AI.
The deployment
In September 2023, at its Connect conference, Meta launched a first series of 28 AI chatbots embodying celebrity-inspired personas, with figures like Kendall Jenner, Charli D'Amelio, Chris Paul or Snoop Dogg lending their image. Meta reportedly paid several million dollars for these deals. Less than a year later, in July 2024, the company quietly pulled these chatbots. None had gathered a notable audience, far behind the personal accounts of the celebrities involved. Meta refocused on AI Studio, a tool that lets users create their own AI characters.
Results Proof C
The Information revealed the withdrawal and several tech outlets (Social Media Today, WinBuzzer) confirmed it, naming the brand, the celebrities and the cause. No detailed financial document, hence C.
How it works
Inferred typical approachThe internal detail is not public. Here is a proven approach that leads to the same result, to adapt to your stack.
The stack in detail
- llm Meta AI (modeles generatifs) Meta's proprietary LLM powering the personas; the exact model is not named in the sources.
- outil Personas conversationnels custom 28 characters built in-house on the image and style of celebrities, under personality-rights agreements.
- plateforme Messenger, Instagram, WhatsApp Distribution channels for the chatbots to users of Meta's apps.
- plateforme Meta AI Studio AI character creation tool for users, which Meta pivoted to after the withdrawal.
Post-mortem
GraveyardWhat happened sourced
In September 2023, Meta presented at Connect a first series of 28 AI chatbots embodying celebrity-inspired personas, with figures like Kendall Jenner, Snoop Dogg, Charli D'Amelio or Chris Paul. According to the press, Meta reportedly paid several million dollars for these deals. The bots did not attract a notable audience. In July 2024, Meta quietly pulled them and refocused on AI Studio, which lets users create their own AI characters.
Reason for failure sourced
The documented cause is the lack of adoption: none of these chatbots gathered a significant audience. Chatting with an AI styled like a celebrity, while the real person is reachable elsewhere, did not find its public. The product rested on a usage bet that did not pan out, despite the investment in the personality deals.
Cost sourced
Several million dollars in celebrity deals, according to the press, for a product pulled in less than a year. Exact amount not published.
Warning signs inferred
Inferred: doubt about the appetite for fake celebrity personas existed from launch. A celebrity's value rests on the authenticity of the connection, hard to replicate with a chatbot. A more modest usage test before the big deals could have revealed the weak traction.
Lessons in hindsight inferred
Inferred: paying personality rights does not create usage. An AI persona does not replace the connection an audience has with a real person. It is better to validate appetite on a lightweight format before committing to costly deals. The pivot to user-created characters suggests the demand was for creation, not for consuming imposed personas.
Inferred: the conversational personas pattern remains valid, but shifted. The failure is about imposed celebrity personas, not about AI agents in general or user-created characters, which Meta pivoted to. The category survives, the celebrity execution failed.
How your customers perceive this type of use
Sourced studiesLes consommateurs n'acceptent pas les chatbots par defaut : 64% prefereraient que les entreprises n'utilisent pas d'IA dans leur service client (Gartner, 2024) et pres d'un utilisateur sur cinq du service client par IA n'en retire aucun benefice (Qualtrics, 2025). L'acceptation se construit sur trois conditions mesurees par Salesforce : savoir qu'on parle a une IA, pouvoir escalader vers un humain, comprendre la logique de l'agent.
Acceptance conditions
- Etre informe qu'on parle a une IA et non a un humain (pres de 75% le demandent, Salesforce 2024)
- Un chemin d'escalade clair vers un agent humain (45% plus enclins a utiliser l'agent IA, Salesforce 2024)
- Une logique de l'agent clairement expliquee (44% plus enclins, Salesforce 2024)
Red lines
- Rendre l'humain injoignable : c'est la premiere inquietude des consommateurs sur l'IA dans le service client (Gartner 2024) et 50% craignent que l'IA les coupe du contact humain (Qualtrics 2025)
- Remplacer le service client par l'IA sans alternative : 53% envisageraient de partir chez un concurrent (Gartner 2024)
Sources: Salesforce 2024 · Gartner 2024 · Qualtrics 2025
How to replicate
Inference, not sourcedData prerequisites
- signal d'engagement par persona
- tests d'usage prealables
Org prerequisites
- accords de droits de personnalite si celebrites
- capacite a arreter vite un format sans traction
Possible stack
- LLM avec personas
- outil de creation de personnages par les utilisateurs
- mesure d'engagement
First step: Tester l'appetence sur un format leger et peu couteux avant tout accord de personnalite, en mesurant l'engagement reel plutot que la curiosite de lancement.
The same trap, elsewhere
Chegg
verticalized AI assistant built as a wrapper on a third-party LLM
DPD
GenAI customer agent
MyCity (Ville de New York)
public-service RAG conversational agent
Sources
- S1 Meta Scraps Celebrity AI Chatbots That Fell Flat With Users Established press archive pending
- S2 Meta Retires Celebrity-Styled AI Bots Secondary archive pending
- S3 Meta Ends Celebrity AI Chatbot Project Amid Low Interest Secondary archive pending
An error, newer info, a source?
This page lives on its accuracy. If a figure has moved, if the deployment has changed, or if you have a higher-quality source, tell us. Every sourced correction is verified before publication.