Italy’s competition regulator has directed Meta to halt a new policy banning rival AI chatbots from running on WhatsApp through its Business API — stepping up a high-stakes fight over how popular messaging platforms police competing services on their dominant networks.
The Italian Competition Authority, or AGCM, introduced interim measures after discovering there was enough initial evidence that Meta’s rule may unfairly tip the field toward its own Meta AI assistant by effectively cutting off distribution for competitors within WhatsApp.
Regulators cautioned that the ban could cause “immediate and irreparable” damage to innovation as well as consumers’ access to the AI assistant market, which is evolving rapidly.
What Italy Ordered and Why the WhatsApp AI Ban Was Halted
The order calls for Meta to suspend its restriction as the watchdog’s investigation proceeds. The question is whether a platform with WhatsApp’s reach can legally prevent general-purpose AI bots developed by third parties from functioning through the same business tools that many developers and companies already use to interact with users.
AGCM said it is looking into whether the policy represents an abuse of dominance by restricting market access and slowing technical development. WhatsApp’s scale, and the strong network effects in messaging, means that the regulator believes no longer having access to that channel can materially impact a rival bot’s chances of competing — certainly not if a helper app is predicated on conversational engagement and rapid feedback loops feeding data back into its learning loop.
How Meta’s Policy Works on WhatsApp Business API
Meta recently adjusted its WhatsApp Business API terms to prohibit distribution of general-purpose AI chatbots (like OpenAI, Anthropic, or Perplexity assistants) via the API. The company says the API is built for customer service and commerce workflows, not as a marketplace for chatbots, and that users can reach third-party assistants on other surfaces.
Important nuance: The restriction doesn’t apply to businesses using AI to take care of their own customer support on WhatsApp. A retailer’s virtual agent answering order questions is fine. What’s been blocked are general-purpose bots that act as standalone assistants within WhatsApp, such as those from startups creating personal helpers, study aids, or travel bots.
The stakes are high, as WhatsApp has more than 2 billion users around the world, according to Meta’s investor disclosures, and has deep penetration in Italy. For a lot of developers, the way to reach users where they already spend their time is through WhatsApp’s Business API — many of these people don’t want to have to persuade recipients to install new apps or try out untested channels.
European Scrutiny Under the DMA and Gatekeeper Rules
Brussels is also watching. The European Commission has launched a separate investigation into whether the WhatsApp policy illegally shuts out third-party AI makers in the European Economic Area. The case cuts to the Digital Markets Act, which places special requirements on identified “gatekeepers” like Meta to prevent self-preferencing and guarantee fair access for business users.
Admittedly, the DMA doesn’t create open chatbot stores on WhatsApp, but it does take gatekeepers to task for actively harming rivals’ capacity to reach customers over their essential interfaces.
If EU enforcers determine the ban hampers interoperation or unfairly benefits Meta’s own assistant, they may try to get quick fixes.
Impact on Developers and Users While Ban Is Suspended
The suspension, for AI developers, is a lifeline. The teams that stand behind assistants like ChatGPT or Claude, or even newer entries including Perplexity and Poe, have been beta testing WhatsApp experiences for a reason: user acquisition is more affordable, and retention is better in chat environments where people already spend time every day. Pulling the plug on WhatsApp distribution would drive up marketing costs and restrict feedback data that drives iteration.
For customers, the difference in practice is one of selection. While the ban is suspended, users in Italy ought to continue to be able to access third-party assistants inside WhatsApp, along with Meta’s own AI functions. Retailers or banks could inject a customer service chatbot if they wish, for the benefit of consumers and their customers, which are not impacted by the policy as well, so long as they remain agnostic to the antitrust outcome.
What Comes Next in Italy’s Case Against Meta’s Policy
Meta can challenge AGCM’s interim order before a court, which is often the next step in Italian competition cases. Concurrently, the authority will keep collecting evidence and, if it finds an abuse, could demand behavioral commitments or monetary penalties. At the EU level, DMA enforcement provides for fines of up to 10% of revenues worldwide if you’re caught misbehaving, rising to a possible 20% for recidivism — applying considerable pressure on platform policies that undermine rivals.
Italy has been particularly aggressive on the AI and platform governance fronts — it previously constrained how certain AI services can operate in the country, thanks to its data protection authority — so the antitrust move is part of a wider pattern. For now, the memo to Meta is clear: Don’t use the power of WhatsApp to wall off competitors in a category where consumer experimentation and interoperability are central to advancement.
Whether this becomes a Europe-wide norm will depend on the Commission’s analysis. But with regulators citing gatekeeper foreclosure as a shared concern, the calculus has tipped in favor of maintaining open access to WhatsApp for rival AI assistants — at least until legal questions are settled.
