For now, Meta will charge developers in Italy, where the country’s competition watchdog asked the company to suspend its policy last December. The company said that the new pricing for non-template responses will begin on February 16. Meta plans to charge $0.0691/ €0.0572 / £0.0498 per message to developers for AI responses. This could result in steep bills for developers if users are exchanging thousands of queries with AI chatbots every day.

Earlier this month, Meta sent notices to developers creating an exemption for Italian phone numbers and allowing AI chatbots to serve those customers. At that time, the company didn’t mention any plans to charge developers.

Currently, WhatsApp already charges companies for using its API for various template responses to customers, which include use cases like marketing, utility, or authentication. This includes messages users receive about payment reminders and shipping updates. 

“Where we are legally required to provide AI chatbots through the WhatsApp business API, we are introducing pricing for the companies that choose to use our platform to provide those services,” a Meta spokesperson told TechCrunch. This could also establish a precedent for other geographies if Meta has to cave in and allow developers to operate their chatbots.

Meta first announced this past October that it would block all third-party AI chatbots from using WhatsApp through its WhatsApp Business API. 

Meta said its systems weren’t designed to handle responses from AI bots and were being strained. 

“The emergence of AI chatbots on our Business API put a strain on our systems that they were not designed to support. This logic assumes WhatsApp is somehow a de facto app store. The route to market for AI companies is the app stores themselves, their websites, and industry partnerships; not the WhatsApp Business Platform,” the company said at that time.

Since then, various regions, including the EU, Italy, and Brazil, have started anticompetitive probes. Brazil’s watchdog initially asked Meta to suspend the policy. However, a court in Brazil sided with Meta last week and overturned the preliminary order blocking the new policy. As a result, the company has asked developers not to provide their AI chatbots to users in Brazil, TechCrunch has learned. 

Since the policy has kicked in, developers are forced to send a pre-defined message to users of their AI chatbot on WhatsApp to redirect them to their site or app. Providers like OpenAI, Perplexity, and Microsoft announced last year that their WhatsApp bots would not work after January 15, urging users to access them on other platforms.