The European Commission has issued its first interim antitrust measure in roughly 17 years, and the subject is not search engines, browsers or app stores. It is AI chatbots. On June 9, the Commission ordered Meta to restore rival AI assistants' access to the WhatsApp Business API within five working days, on the same terms that applied before Meta shut them out last October. Failure to comply could cost Meta up to 10 percent of its global turnover. Meta called the order regulatory overreach and says it will appeal, with a spokesperson complaining that Brussels has decided "OpenAI and some of the largest companies in the world can use the paid-for WhatsApp Business product for free."
Interim measures are the emergency brake of European competition law: a way to freeze the field before a years-long investigation concludes, used so rarely that the last one predates the smartphone era as we know it. Deploying one here says something specific about how regulators now see AI assistants. The complaint came from three smaller players, including The Interaction Company and the French startup Agentik, who argued that losing WhatsApp access could kill them long before any final ruling. Competition commissioner Teresa Ribera made the logic explicit: "In rapidly evolving markets, competition can be lost long before a final decision is adopted." The order stands while the investigation runs, potentially until 2029.
The underlying fight is about distribution. WhatsApp is where billions of people already are, and an AI assistant that can live inside it inherits that reach overnight. Meta, which has its own assistant ambitions, changed its business API terms in October 2025 to bar external chatbots, a move rivals read as the platform pulling up the ladder just as the market formed. The Commission's intervention treats access to a dominant messaging platform the way earlier eras treated access to the telephone network: infrastructure first, product second. For anyone building on someone else's platform, the lesson cuts both ways. Regulators may now move fast enough to matter, but your business model can still hinge on a five-day order from Brussels.
The timing gives the order extra weight, because Europe's main AI law is about to grow teeth of its own. On August 2, the EU AI Act becomes fully applicable, bringing its high-risk regime into force for companies using AI in hiring, healthcare, education, credit and critical infrastructure. Penalties for the most serious violations reach 35 million euros or 7 percent of global turnover. Law firms across the continent are sending out the same checklist: inventory every AI tool in the building, classify the risk, document who oversees it. The most common client mistake, one Spanish firm notes, is assuming that "just testing" a tool creates no legal exposure.
Put the two stories together and the shape of the season is clear. For three years, Europe's role in AI has mostly been drafting: frameworks, timelines, tiers of risk. This month it started enforcing, and it chose to begin not with a fine for past conduct but with an order shaping who gets to compete right now. Washington spent the same week promising the industry a voluntary framework. Brussels handed it a deadline.