Clones Flopped: Meta Quietly Buries Its AI Celebrities

Meta has officially wound down its project to build AI avatars imitating global stars. What was pitched a year ago as the future of social networks turned out to be an expensive toy nobody wanted. Users did not warm to the plastic conversations with digital copies of Snoop Dogg and Tom Brady, preferring more flexible and less "censored" alternatives. The company is now shifting its attention to the AI Studio platform, which lets any user create their own bots. It is a significant admission: even huge budgets and A-list faces do not guarantee success if the technology does not solve a real problem for the user. *Meta has been designated an extremist organization and is banned in Russia.

AI-processed from Futurism; edited by Hamidun News
Clones Flopped: Meta Quietly Buries Its AI Celebrities
Source: Futurism. Collage: Hamidun News.
Listen to article

Imagine deciding to spend tens of millions of dollars so your users can chat with an AI version of Snoop Dogg who teaches them to play Dungeons & Dragons. It sounds like a plot from a satirical series about Silicon Valley, but for Mark Zuckerberg it was reality just a year ago. Meta's ambitious experiment of putting celebrity "personalities" into its messengers has now ended in a quiet admission of defeat. The company has removed all AI character accounts based on real people from Instagram, WhatsApp and Messenger, effectively acknowledging that the project never took off.

When Meta announced the project at the Connect conference in 2023, the industry read it as an attempt to grab the entertainment chatbot market, where the startup Character.ai was already dominant. Zuckerberg signed a whole constellation of stars, from Kendall Jenner and Naomi Osaka to Tom Brady and Paris Hilton. Each of them was reportedly paid a fee of several million dollars for the right to use their likeness and voice to train the models. The assumption was that these celebrities' armies of fans would deliver explosive growth in messenger activity. But reality turned out to be far more prosaic than the marketing plans and polished presentations.

The core problem was that the bots were plainly boring. Behind the faces of global stars sat the then fairly limited Llama 2 model, wrapped in rigid corporate safety filters. Instead of lively, cheeky conversation, users got sterile, predictable replies that only vaguely resembled the originals' manner of speaking. On top of that, Meta for some reason gave them fictional names. Snoop Dogg became "Dungeon Master" and Kendall Jenner became "Billie". That created an extra cognitive barrier: users had to get used to new character names that looked like their idols but behaved like polite robot consultants from a bank's tech support.

The quiet shutdown says a lot. First, it concedes that "stardust" does not work in generative AI as effectively as it does in traditional ads for sneakers or perfume. People come to chatbots for utility or for sincere, even if simulated, empathy. Meta's project offered neither. Users quickly saw through the fake and went back to ordinary chats. Second, it signals a sharp change of strategy inside the company. Meta is now betting on AI Studio, a platform where any content creator or ordinary user can build their own bot. It is a move from a centralized model of "we give you ready-made idols" to one of "create whatever you like yourself".

It is interesting to watch how quickly Meta cuts costs and admits mistakes when hyped technologies are involved. A similar fate previously befell NFTs on Instagram and many ambitious "metaverse" projects. Now digital clones have joined that list. It is an important lesson for the whole industry: even with access to A-list faces, unlimited server capacity and the world's best engineers, you cannot make people talk to something that does not genuinely interest them. Technology for technology's sake has once again lost to common sense and the real needs of the audience.

What does this mean for the future of AI in social networks? Meta is clearly trying to find a format in which algorithms become an organic part of conversation rather than a sticker on the interface. The company now believes the future belongs to millions of small, highly specialized bots made by users themselves, not to a dozen expensive digital mannequins. It is a more democratic but also more chaotic path. The success of competitor Character.ai has shown that users value context, role-play and the absence of rigid limits more than an official license from a Hollywood star's agent. Zuckerberg had to learn that lesson at a cost of tens of millions of dollars.

Bottom line: The era of expensive corporate AI toys is ending. Users chose functionality and flexibility over paid celebrity marketing. Can Meta catch up with the market through its bot builder, or is it already too late?

*Meta has been designated an extremist organization and is banned in Russia.

Story
ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Unified AI Workspace · Nomad

Test Frontier LLMs & Autonomous Agents in One Workspace

Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.

Try Nomad AI →Launch Nomad Agent
What do you think?
Loading comments…