The Italian Competition Authority Imposes Interim Measures in its Abuse of Dominance Case Against Meta AI

As previously covered on this blog, Meta recently decided to integrate its AI assistant (Meta AI) into its WhatsApp platform, introducing the ability to chat to Meta AI at various points within the WhatsApp app. Its conduct raised eyebrows as a potential abuse of dominance, and later became the subject of an investigation by the Italian Competition Authority (the AGCM). This blog post covers two subsequent developments of that case, Meta’s decision to double down on its behaviour by updating its terms to stop third parties offering rival AI assistants on its platform, as well as the AGCM’s subsequent decision to expand the scope of its investigation and impose interim measures aiming to prevent further harm to competition.

To recap: the AGCM’s investigation

On the 30th of July 2025, the Italian Competition Authority (AGCM) filed Case A576 against Meta for deciding “to pre-install its artificial intelligence service [(Meta AI)] on the WhatsApp app”. As detailed in a previous post, Meta’s decision to integrate its AI assistant into WhatsApp was a fairly clear cut abuse of dominance case, so the AGCM’s decision was relatively unsurprising. To briefly recap, AGCM alleges that WhatsApp is dominant in the relevant market for “consumer communications apps”, and Meta has imposed “the use of its chatbot and AI assistance services on its users […] without any prior request from users". In other words, the Authority is concerned that Meta has tied its AI assistant into WhatsApp. This conduct, in the eyes of the AGCM, allows Meta to channel “its customer base into the emerging market [for AI assistants], not through merit-based competition, but by ‘imposing’ the availability of the two distinct services upon users, potentially harming competitors.”

Meta doubled down
As observers had noticed, Meta also offers a WhatsApp Business Solution alongside its consumer-facing platform, which allows third-party undertakings to communicate with customers by messaging them directly on WhatsApp. This fact is relevant to the case at hand because some third-party undertakings have launched their AI assistants on WhatsApp using the Business Solution, including Open AI’s ChatGPT. In a manner reminiscent of its “move fast and break things” era, Meta has doubled down on its apparently abusive behaviour by updating its WhatsApp Business Solution Terms to strictly prohibit “[p]roviders and developers of artificial intelligence or machine learning technologies, including but not limited to large language models, generative artificial intelligence platforms, general-purpose artificial intelligence assistants, or similar technologies” from using its Business Solution. This change comes into effect on January 15 2026, after which OpenAI has indicated that “ChatGPT will no longer be available on WhatsApp”.

Such a move by Meta was easily foreseen, not least because Meta has a clear incentive to undermine its rivals in the competitive market for AI assistants. Yet as I noted in my original blog post on this topic, and in a subsequent follow-up paper, the recent Android Auto judgement would prohibit Meta from preventing third-party undertakings from providing a rival AI assistant through its WhatsApp Business Solution, since its Solution was clearly built “with a view to enabling third-party undertakings to use it” (para 47). As such, Meta’s new terms of conditions likely constitute an infringement competition law in and of themselves.

Reaction of the AGCM

Meta’s decision to actively interfere with the ability of rivals to compete on its dominant platform therefore constitutes a refusal to deal under the Android Auto doctrine. This fact did not escape notice of the AGCM. On the 25th of November 2025, the Authority announced that it had “expanded [its] investigation procedure” against Meta to include the aforementioned WhatsApp Business Solution Terms.

More importantly, the AGCM also decided to initiate “the procedure for the adoption [of] precautionary measures”, i.e. interim measures, regarding Meta’s integration of its AI assistant into WhatsApp and its updated Terms. The decision to pursue interim remedies is welcome, and one which I have repeatedly suggested in the face of such a clear abuse and its concomitant ongoing harm to competition. Especially following the recent ban on competing AI assistants being offered through WhatsApp, reporting indicates that rivals have also requested interim measures. The AGCM’s recent case update acknowledged the urgent need for intervention, pointing out that the “unrepeatable” opportunities to train AI models in the current moment, and the “exponential” growth of AI assistant services, and the fact that Meta has “exclusive” access to an “enormous” user base, means that the undertaking’s conduct may “may seriously and irreparably affect market” outcomes.

In terms of next steps, Meta now has a week to submit arguments defending itself against the imposition of interim measures. Paragraph 14 of the decision explains that if the interim measures are subsequently implemented, they will have two overarching effects. First, they will suspend the application of the new WhatsApp Business Solution Terms, thereby allowing third-party undertakings to launch rival AI assistants on WhatsApp. Second, they will broadly prohibit Meta from making further modifications to WhatsApp which would “amplify the presence of Meta AI on WhatsApp to the detriment of competitors”. Whatever happens, given the pace of this case and the apparent unwillingness of Meta to choose a business strategy that is commensurate with EU competition law, it seems that this case will continue to be one to watch over the coming months and potentially years.

***

N.b. English translations of the original Italian decision were provided by the author through an automatic translation tool.