Introduction

With the latest intervention by the European Commission (‘EC’) and the Italian Competition Authority (‘AGCM’) in the WhatsApp Business API case, the question surrounding the nature of WhatsApp’s services has become increasingly critical. Competition authorities across the world raise eyebrows when dominant platforms attempt to block third-party service providers, observing that such access often constitutes an indispensable gateway between businesses and consumers. This regulatory wave leads to a compelling debate: Has WhatsApp become so central to digital life that its API should be treated as an essential facility?

Key Interventions by Italian AGCM and the EC Against the WhatsApp Business API<

In a major regulatory crackdown, the Italian AGCM, in December 2025, took the radical step of ordering Meta to suspend its new terms for the WhatsApp Business API. The AGCM observed that these terms, which restrict access for third-party AI chatbots, likely constitute an abuse of dominance under Article 102 of the Treaty on the Functioning of the European Union (‘TFEU’). The introduction of these restrictive terms reflects Meta’s leveraging of its dominance in the app-based messaging market to stifle competition in the AI chatbot services market.

The AGCM’s action aligns with a broader inquiry by the EC, which, earlier in the same month, launched a formal investigation to determine whether the same terms breach European Union (‘EU’) competition law by preventing AI providers from operating via WhatsApp. These actions stem specifically from concerns that AI providers are prevented from competing fairly when their access to the platform is blocked. Access is vital for reaching customers and utilizing Business Solution data for training and development purposes. Interestingly, despite these new terms, “Meta AI”, which is Meta’s own chatbot integration, remains fully accessible through the platform.

Since the WhatsApp Business API is the foundation for creating value-added services, enhancing customer reach and improving customer engagement, restricting third-party AI chatbot access could be seen as limiting an essential facility. The key question, however, is whether such access passes the legal threshold of the Essential Facilities Doctrine (‘EFD’).

Testing the WhatsApp Business API Against the High Bar of Essentiality

In the EU, the Bronner test remains a landmark legal standard in competition law for determining when a dominant entity’s refusal to grant access to a facility constitutes an abuse of dominance. The test originated in Oscar Bronner v. Mediaprint, decided by the European Court of Justice (‘ECJ’) in 1998. For a refusal to be considered anticompetitive, three cumulative factors must be met: indispensability, elimination of competition, and a lack of objective justification.

Applying the strict Bronner standard to the WhatsApp Business API reveals a complex tug-of-war. Under a strict interpretation of the indispensability requirement, the facility concerned must be essential for the AI service providers’ business, with no actual or potential substitutes available. Pursuant to this interpretation, Meta may argue that the platform is not indispensable because AI providers currently can reach customers via Telegram, Signal, or standalone apps.

However, following a series of EU decisions that revolutionised the concept of indispensability, the bar has been lowered, shifting instead toward commercial viability. Notably, the 2007 Microsoft ruling by the European Court of First Instance (‘CFI’) examined the “network effects” of Windows and recognised that if a competitor’s server could not interoperate with the Windows PCs, it would be commercially dead on arrival. Today, Meta may argue that AI providers do not need WhatsApp, as they can use Telegram or standalone apps. But under the Microsoft precedent, regulators may fire back because over 2 billion users live on WhatsApp, and it is economically unviable for an AI provider to reach such a vast user base elsewhere.

Further, the recent Google Android Auto case shares a similar factual matrix to that of WhatsApp Business API case. Both involve an abuse of dominance through a refusal of interoperability. This case represents a pivotal precedent that fundamentally lowers the threshold for holding digital platforms accountable, specifically those already open to third parties. Because the Android Auto platform was already open to third parties, the ECJ ruled that it could not rely on the “shield” of the strict Bronner requirement of indispensability. In such cases, competition authorities only need to show that the refusal is likely to result in anticompetitive effects.

Similarly, Meta has already opened the WhatsApp Business API to an ecosystem of providers, the AGCM may not be required to prove that the WhatsApp Business API constitutes an “essential facility”. Instead they may only need to prove that the new terms by effectively refusing to grant access are capable of producing anticompetitive effects.

Furthermore, just as Google used its control in the Google Android Auto case to influence the market for charging stations, Meta currently leverages its control over WhatsApp to influence the business software market, specifically the AI chatbot services. The API acts as a bottleneck, allowing Meta to leverage its dominance in the app-based messaging into the AI chatbot services. By banning third-party AI chatbots while simultaneously integrating “Meta AI” into the chat interface, Meta arguably engages in self-preferencing.

While the broad AI market remains competitive, Meta’s restrictive terms allow it to secure a monopoly within the “Conversational AI-in-Messaging” niche. The direct effect of the ban is the elimination of choice for WhatsApp’s 3.3 billion monthly active users, as Meta ensures its native AI is the only open-domain assistant available within the platform. Consequently, third-party AI chatbots are denied access to the WhatsApp messaging interface, massive user base and interaction layer, all of which are essential for competing effectively in the AI chatbot services market.

Earlier cases, such as Bronner and IMS Health, required the elimination of all competition in the relevant market. Conversely, the Microsoft ruling clarified that this standard refers to the elimination of “effective” competition. This means that a dominant entity need not drive every single competitor to zero revenue, rather it is sufficient to make it nearly impossible for competitors to pose a meaningful competitive restraint on its power.

Further, the CFI in the Microsoft case expanded the Magill/IMS Health “new product” requirement for interoperability refusals, adopting harm to technical development instead. This development is important as earlier refusal was considered abusive only if it prevented the emergence of a new product. Post-Microsoft, the refusal is abusive if it forecloses technical development within existing markets. This argument could be directly used by the EC and AGCM in the given case stating that Meta’s refusal to share its essential API amounts to limiting technical development.

Meta has presented a defense rooted in “infrastructure strain”, arguing that general-purpose AI chatbots represent an unprecedented use case that places enormous pressure on the WhatsApp systems in the ongoing case. According to the company, the high volume of media uploads, open-ended messaging patterns, and voice interactions required by these AI assistants far exceed the intended design of an API originally built for predictable business-to-customer support. This argument may be insufficient in light of the Google Android Auto case where ECJ held that a dominant entity operating a platform originally designed for third-party use can only refuse access based on genuine objective justifications, namely, technical impossibility or substantiated security or integrity concerns.

The reasoning by Meta is unconvincing from another perspective. In the Microsoft case, The EC observed that a dominant entity abuses its power when its refusal to supply prevents the emergence of a new product for which there is potential customer demand. By imposing the ban, Meta is not only protecting its API but also foreclosing an entire category of innovative services. The claim on volume is self-contradictory since the company integrates its own AI into the same infrastructure which is calls as a capacity burden.

Conclusion

Placing reliance on important precedents regarding the EFD, the recent tussle over the restrictive terms for WhatsApp Business API may well satisfy the evolving test of essentiality. Over the years, the high bar of essentiality has been lowered by national and European courts in the EU and there lies a possibility that the API will be considered essential enough for third-party AI chatbots to effectively compete and pose a meaningful competitive restraint on Meta’s dominance. The AGCM’s recent interim measures, which include the suspension of Meta’s ban on third-party AI chatbots, demonstrate a willingness to enforce DMA-style openness through traditional antitrust tools.

Soujanya Boxy is a law graduate from National Law University, Odisha.