Meta to ban AI chatbots like Perplexity AI from WhatsApp:The company orders third-party AI assistants to shut down by January 15, 2026

Until recently, we were seeing how Perplexity AI and ChatGPT could be used directly within WhatsApp, but that’s no longer the case.
WhatsApp is now drawing a clear line between chatting with humans and chatting with machines, as it moves to ban general-purpose AI chatbots from the popular messaging platform. Meta, the company behind WhatsApp, has quietly updated its Business API rules to ban all general-purpose AI assistants from operating on the platform starting January 15, 2026. That means AI chatbots like OpenAI’s ChatGPT, Perplexity AI, Luzia, and Poke, which millions have been using through WhatsApp, will soon have to pack up and leave. What’s changing exactly Meta has added a new section to WhatsApp’s Business API policy that directly targets AI-based services. The updated rule says: AI providers offering general-purpose assistants are strictly prohibited from using WhatsApp’s Business API. In simple terms, if your WhatsApp bot mainly acts as an AI assistant, it’s not allowed anymore. But not all bots are banned Meta clarified that this ban doesn’t apply to customer-support bots or businesses that use AI in limited ways. For instance, an airline chatbot that helps you check flight status or a hotel assistant that allows you to modify bookings can continue as usual. Why Meta is doing this Meta says WhatsApp’s Business API was never meant to be a hosting ground for general AI chatbots. It was created for companies to communicate with their customers, not to run massive AI services. Over the past year, Meta has noticed a surge in developers using the API to build AI chat assistants that handle millions of messages daily. These interactions, from voice notes to image uploads, put heavy pressure on WhatsApp’s infrastructure, something Meta didn’t plan for. The company hinted that these AI tools needed “a different kind of support”, suggesting that WhatsApp’s systems, pricing, and moderation models aren’t built to handle the scale or complexity of such assistants. The money factor There’s also a clear business side to this decision. WhatsApp’s Business API is one of Meta’s biggest money-makers; companies are charged per message for categories like marketing, customer service, or authentication. But for AI bots like ChatGPT or Perplexity, there wasn’t a pricing system in place. So, while these companies ran massive chat operations on WhatsApp, Meta earned little to nothing from the surge in traffic.
By banning these bots, Meta ensures it keeps full control over how WhatsApp is used and monetised. Meta AI stays, others go Interestingly, this decision leaves only one AI assistant standing, Meta’s own chatbot, Meta AI. It will continue to live on WhatsApp, while third-party chatbots disappear. So, from early 2026, users won’t be able to chat with external AI bots for quick answers, image generation, or document analysis. WhatsApp will go back to being what Meta wants it to be, a business communication platform, not an AI playground. The end of the AI rush on WhatsApp Over the past year, AI startups have rushed to bring their assistants to WhatsApp. OpenAI’s integration lets users chat, share images for analysis, and even ask questions about PDFs directly inside the app. Perplexity AI’s bot offered real-time answers and summaries, making WhatsApp feel like an AI super-app. But those experiments won’t last long. With the new rules, Meta has made it clear: WhatsApp is for customer-business conversations, not for AI companions.