COMMENTARY: Meta versus Brussels over WhatsApp — the duel redefining the AI market

COMMENTARY: Meta versus Brussels over WhatsApp — the duel redefining the AI market

Introduction: An emergency order unprecedented since 2009

June 9 — the Commission pulls out the heavy artillery

On June 9, 2026, the European Commission adopted a historic interim measures decision: it ordered Meta to restore free access to WhatsApp for competing AI assistants within five business days. This was only the second time — since the Broadcom case in 2009 — that the Commission deployed this exceptional tool in European competition law. The message from Executive Vice President Teresa Ribera was unambiguous: "In rapidly evolving markets, competition can be lost well before a final decision is adopted."

The decision aimed to end a sequence that began on October 15, 2025, when Meta modified its terms of use for the WhatsApp Business API to exclude all third-party AI assistants — leaving only its own tool, Meta AI, accessible. Competitors including ChatGPT from OpenAI, Claude from Anthropic, Perplexity, Luzia, Poke, and others had been banned from the world's largest messaging service. Then, in March 2026, Meta appeared to reopen — but by charging a per-message fee the Commission deemed "equivalent in practice to the initial ban."

What is at stake

To understand the issue, one must understand WhatsApp's place in the digital lives of 2 billion users. This is not just another app. It is the daily communication interface for a large share of the planet. In Europe particularly, WhatsApp is the dominant messaging channel. Controlling it means controlling the entry point for AI assistants reaching consumers. That is precisely what the Commission identified as the central competitive risk: if only Meta AI can access this channel, competition in the emerging general AI assistant market could be permanently locked before the rule of law is even established.

The interim measures will apply until June 2029 or until the conclusion of the main antitrust investigation — whichever is later. During this period, Meta must maintain access terms in place before October 15, 2025 — meaning free access for all general AI assistants. Non-compliance exposes Meta to fines of up to 10% of its global annual turnover — approximately 18 billion euros by industry estimates.

The chronology of the standoff

October 2025: Meta's power play

It all started on October 15, 2025: Meta modified its WhatsApp Business API terms of use to ban general AI assistants — any tool whose primary function is as an AI chatbot — from accessing the service. The change took effect on January 15, 2026 for existing providers, and immediately from October for new entrants. In practice, only Meta AI — the in-house chatbot — remained accessible on the platform.

The first complaints arrived within weeks. The Interaction Company, French startup Agentik, and a competing Spanish company filed complaints with the European Commission. The Commission opened a formal investigation in December 2025. In February 2026, it notified Meta of its "preliminary assessment": the policy appeared, prima facie, in violation of EU competition rules, specifically Article 102 TFEU prohibiting abuse of dominant position. The Commission expressed its intent to impose interim measures.

March 2026: Meta feigns a concession

On March 5, 2026, in an attempt to defuse the crisis, Meta announced it would readmit competing AI assistants to WhatsApp via its API — but for a fee. The pricing ranged from €0.049 to €0.1323 per "non-template message," depending on the country. For AI startups whose business model depends on massive, free interactions, this cost was prohibitive. Meta had found an elegant way to maintain de facto exclusion while claiming to open up.

The Commission was not fooled. It continued its investigation, stating it was "analyzing the impact" of this change. In April 2026, it issued a supplementary statement of objections confirming its intent to order interim measures. And on June 9, it acted. The verdict was clear: Meta's fee was "equivalent in practice" to the original ban. Meta was required to return to pre-October 2025 terms — full free access.

Meta's legal position and its appeal

Meta says no

Meta called the measure "regulatory overreach" and announced its intention to appeal before the Court of Justice of the European Union. The company's argument rests on several pillars: first, that it does not hold a dominant position in the general AI assistant market; second, that the WhatsApp Business API is not a "crucial channel" for distributing AI chatbots, since these can be distributed through app stores, websites, and industry partnerships; and third, that regulatory intervention is premature in a market still taking shape.

These arguments are not without foundation. European case law on mandatory access to essential infrastructure — the "essential facility" doctrine — imposes strict conditions regulators must meet. Meta will argue that WhatsApp does not qualify as an essential facility in the legal sense. This is a genuine legal debate that will take years to resolve before the courts.

The Broadcom precedent and its limits

The Broadcom case of 2009 — the only precedent for interim measures under Regulation 1/2003 — had forced the semiconductor manufacturer to cease certain product bundling practices. That precedent had been judged substantively favorable to the Commission. But the Meta situation is more complex: it involves an emerging digital market, a tool — AI — whose regulatory contours remain unclear, and a company whose grip on daily communication is without historical precedent.

The law firm Houthoff, analyzing the decision, notes that interim measures are "used with caution" by the Commission, and that the threshold — prima facie evidence of an infringement AND risk of serious and irreparable harm to competition — is high. The Commission nonetheless cleared it. This suggests its case is solid — but Meta's appeal is not without merit.

Who will win? The possible scenarios

Scenario 1: The Commission holds

If the EU Court rejects Meta's appeal — as it did in the Broadcom case — the interim measures remain in place until June 2029. The main substantive investigation continues. If it finds an abuse of dominant position, Meta faces a fine of up to 10% of global turnover and permanent structural remedies. This scenario is favorable to competition in the European AI assistant market.

Under this scenario, players like OpenAI, Anthropic, and European AI startups would benefit from free access to WhatsApp for several years — enough time to establish themselves as credible alternatives to Meta AI. The Commission would have achieved what it aimed for: preserving market contestability in its formative phase.

Scenario 2: Meta wins on appeal

If the Court suspends the interim measures — as it did in the IMS Health case in 2001 — Meta regains its pricing freedom. Competing AI products once again face a financial wall on WhatsApp. Competition on this channel is frozen. The Commission must wait for its final decision — years away — to impose remedies.

This is the scenario European AI startups most fear. Three years is an eternity in the AI sector. Markets forming today will be crystallized tomorrow. If Meta AI benefits from a de facto exclusivity on WhatsApp during this period, future competition becomes structurally difficult. That is precisely the Commission's argument: the harm will be "nearly impossible to repair" if not prevented now.

The DMA lying in wait

Why the Digital Markets Act falls short

A legitimate question arises: why is the Commission resorting to traditional competition law — Article 102 TFEU — rather than the Digital Markets Act (DMA), specifically designed to regulate digital gatekeepers? The answer is technical but important: general AI assistants are not explicitly listed among the "core platform services" subject to the DMA. AI is not — yet — directly covered.

This gap was acknowledged by the Commission in its April 2026 DMA review report, which identifies AI and cloud as areas "requiring particular attention." Pending a DMA revision, Article 102 remains the available tool. This situation illustrates the recurring reality of digital regulation: technologies evolve faster than legislation. The Commission improvises — not always badly, but improvises nonetheless.

Meta's partial court victory

Paradoxically, at the very moment the Commission was tightening the screws on WhatsApp, the EU Court handed Meta a partial victory on another front: on June 3, 2026, it annulled the designation of Facebook Marketplace as a core platform service under the DMA, finding that the Commission had used "a presumptive and incomplete analysis." However, it maintained the designation of Messenger. This mixed result illustrates the complexity of the ongoing legal battle and Meta's real capacity to successfully challenge European regulatory decisions.

This context is important for evaluating Meta's chances in its appeal against the WhatsApp AI interim measures. The company is not without legal resources, nor without successes. It has proven it can win against Brussels. The Commission knows this. That is why it built its case on solid foundations — but nothing is guaranteed in proceedings this novel.

The implications for Europe's AI ecosystem

An opportunity for European players

If the interim measures hold, they create a real window of opportunity for European AI players. Companies like French startup Mistral AI, tools like Perplexity, or specialized AI assistants will be able to access WhatsApp's user base in Europe without paying a toll to Meta. This is not a guarantee of success — product quality remains the determining factor — but it is a leveling of the playing field at a critical moment.

The strategic issue is real: who controls the AI assistant that Europeans use daily? If it is exclusively Meta AI, European consumers' dependence on an American platform is total and structural. If multiple competing assistants coexist — including European players — Europe's digital sovereignty in AI is better preserved. The Commission has perfectly identified this issue. Its decision is as much geopolitical as it is antitrust.

A global signal

The European decision has global resonances. Brazil adopted a similar temporary injunction against Meta in January 2026 — Meta obtained a suspension of that injunction on appeal, but proceedings are ongoing. COMESA — the Common Market for Eastern and Southern Africa — received a similar complaint. Italy opened its own proceedings. The question of open access to messaging platforms for competing AI assistants is now a global regulatory issue.

Meta faces an increasingly hostile regulatory environment on this question. Even if it wins its Brussels appeal, it risks unfavorable decisions in other jurisdictions. And the reputational costs — being seen as a company deliberately blocking AI competition — are not negligible at a time when user and regulator trust is a strategic asset.

European AI startups: a concrete window of opportunity

Mistral, Aleph Alpha, and the direct beneficiaries

If the interim measures hold, it is European artificial intelligence startups that will benefit most directly. Mistral AI, the Paris-based gem valued at several billion euros, develops language models competitive with OpenAI's. Being able to distribute its services via the WhatsApp API at no cost — without paying a royalty to Meta — means accessing a user base of 2 billion people without prohibitive distribution costs. This is exactly the kind of market equalization the Commission's decision seeks to produce.

Aleph Alpha, the German startup specializing in AI for enterprises and governments, Axel Springer AI, Poe, and other European players could also benefit. The stakes go beyond the consumer market: companies deploying AI assistants for their clients via WhatsApp Business represent a substantial market across Europe, particularly in banking, insurance, and customer service. If Meta reasserts control over this channel, an entire ecosystem of enterprise applications will be penalized.

The paradox of the European champion

There is a paradox in this situation: Europe, which struggles to create world-scale technology champions, finds itself defending its startups' access to an American company's infrastructure. The dependency is real: without WhatsApp, European AI startups must build their own distribution channels — mobile apps, websites, direct integrations. This is not impossible, but it is costly and slower to deploy at scale.

That is precisely why the Commission's decision goes beyond classical antitrust. It touches on digital sovereignty: can Europe build its AI ecosystem if distribution channels are controlled by American actors who can alter the rules at will? The answer lies in regulation — as the Commission is doing — and in investment in European alternatives. Both are necessary.

Meta's response: between denial and adaptation

Silicon Valley's adaptation strategy

Meta's initial reaction to the European decision reflects a well-worn strategy by major American tech companies facing regulators: first denounce "regulatory overreach," then adapt the business model while minimizing real concessions. The per-message fee proposed in March 2026 — between €0.049 and €0.1323 — was Meta's version of an "opening" that maintained economic exclusion. The Commission saw through that game.

The real question is whether Meta will choose, after exhausting its legal remedies, to durably adapt to the European regulatory environment or restrict its services in Europe as it has done in other domains. Europe represents a market of 450 million consumers that Meta cannot afford to ignore. A complete withdrawal is virtually impossible. A strategic adaptation — seeking to monetize in other ways what the Commission forced open — is far more likely.

Meta AI facing the obligation of excellence

If Meta AI must coexist with ChatGPT, Claude, and European assistants on WhatsApp, it will need to compete on quality rather than channel exclusivity. Paradoxically, this is good news for WhatsApp users: competition drives improvement. If Meta AI is the best tool, users will choose it. If Anthropic's Claude or Mistral delivers better results, users will migrate. That is how a competitive market normally functions — exactly what the Commission seeks to preserve.

There is an irony in this situation: the regulatory decision forces Meta to play the meritocracy game on its own platform. Meta AI, with Meta's enormous resources — billions invested in Llama models, computing infrastructure, and research teams — is not without assets. But the artificial distribution advantage had removed the necessity for excellence. European regulation reintroduces that necessity.

Conclusion: The AI market cannot belong to a single platform

A principle worth defending

The standoff between the European Commission and Meta over WhatsApp raises a question of principle that goes far beyond the arcana of competition law: can dominant platforms use their position to lock up adjacent markets in favor of their own products? Europe's answer is no. And it has the legal instruments — interim measures, DMA, Article 102 — to enforce that no.

What is playing out on WhatsApp in 2026 is the same question that arose over web browsers with Microsoft and Internet Explorer in 2001: can a dominant actor impose its own product by excluding competitors from its infrastructure? Europe said no to Microsoft. It is saying no to Meta. The consistency is there. The question is whether it has the tenacity to follow through.

Who will win?

My honest answer: I don't know. The interim measures create a more open market reality for three years. If the main investigation confirms the abuse, Meta will pay a historic fine and will need to open its infrastructure permanently. If it is overturned or if courts side with Meta on appeal, the AI market on WhatsApp will be locked. In either case, this fight was necessary. Europe had a duty to act. History will judge whether it acted quickly and firmly enough.

By Maxime Marquette, columnist

Columnist's transparency note

My declared biases

I favor proactive regulation of large technology platforms, and I consider that digital market concentration poses a risk to democracy and innovation. I do not personally use Meta AI as my primary tool and have no financial conflicts of interest with the parties involved. My analysis is based on public legal sources, specialized media reports, and official European Commission communications.

What I don't know

I am not a specialist in digital competition law. I cannot evaluate the precise strength of Meta's appeal arguments nor predict the outcome of judicial proceedings. The outcome of the interim measures and the main investigation remains uncertain, as I indicated in my analysis.

Sources

Primary sources

European Commission — Interim measures decision against Meta, preserving WhatsApp access — June 9, 2026

Houthoff — A rare show of force: Commission deploys interim measures against Meta — June 23, 2026

BBC News — EU orders Meta to open WhatsApp to rival AI chatbots — June 9, 2026

European Commission — Ribera remarks on interim measures against Meta — June 9, 2026

Secondary sources

TechCrunch — Meta will allow rival AI chatbots on WhatsApp in Europe, but for a fee — March 5, 2026

CNBC — EU signals intent to impose interim measures on Meta over WhatsApp AI policy — February 9, 2026

Tech Xplore — EU court gives Meta partial win on tech rules — June 3, 2026