Meta-owned WhatsApp is making significant changes to its platform’s policies and features as it targets restricting general-purpose AI assistants and combating message spam.
WhatsApp has updated its Business API policy to ban general-purpose chatbots from its platform, a move that will likely affect popular third-party AI assistants like OpenAI’s ChatGPT, Perplexity, Luzia, and Poke.
The new terms, which take effect on January 15, 2026, prohibit “AI Providers” from using the WhatsApp Business Solution to distribute their AI assistants when the technology is the primary functionality offered.
Meta stated that the WhatsApp Business API was designed for business-to-customer (B2C) customer service and relevant updates, not as a distribution platform for general chatbots. The unanticipated use of the API by these AI services placed a heavy burden on WhatsApp’s system due to increased message volume and a need for different support.
This ban does not affect businesses that use AI for customer service purposes—for example, a travel company using a bot to support its customers. The focus is on use cases outside the API’s “intended design and strategic focus.”
The change effectively makes WhatsApp unavailable as a platform for distributing AI solutions like general assistants or agents, leaving Meta AI as the only available assistant on the chat app.
A key factor is likely revenue. WhatsApp charges businesses for different message templates (marketing, utility, support), and the general-purpose chatbots did not fit within this structure, preventing the company from charging for their high message volume usage. Meta CEO Mark Zuckerberg has previously emphasized business messaging as a major future revenue pillar.
New Limits to Fight Spam
In a separate move to address a growing spam problem, WhatsApp is testing a new measure that curbs the number of messages both individual users and businesses can send to unknown people without getting a response.
How it Works: All messages sent to a contact who has not yet replied will count against a per-month limit. For example, a few initial messages sent to a new acquaintance before they respond would count toward the cap.
The control is specifically designed to be effective against users and businesses that “blast messages and spam people,” with the company expecting average users not to be affected.
WhatsApp is currently testing different limits and will display a warning pop-up when a user or business is close to hitting the cap to prevent them from being blocked from sending messages.
This move is the latest in a series of steps WhatsApp has taken to reduce spam and improve the user experience, following recent tests like limiting the number of marketing messages businesses can send and providing users with an option to unsubscribe from marketing messages. The company is expanding this experiment in over a dozen countries, including its massive market of India.