When ChatGPT vanished from WhatsApp in mid-January, 50 million users lost access overnight. So did businesses that had spent months building customer service workflows around third-party AI chatbots on the platform.
Meta’s explanation was straightforward: the WhatsApp Business API wasn’t designed for standalone AI assistants. The company was simply enforcing updated terms it had announced in October. The only AI that stayed was Meta’s own.
It looked like a routine platform policy change. Brussels saw it differently.
Two months after opening an investigation, the European Commission sent Meta a Statement of Objections and threatened emergency action. The speed alone is unusual. But what matters more is the question regulators are forcing into the open: when a communications platform embeds its own AI, does it have to let competitors in too?
That question doesn’t stop at WhatsApp. Every major UC and collaboration vendor is making similar choices right now about how open their AI layer will be.
Brussels moves at Silicon Valley speed
The European Commission doesn’t usually work this fast. Antitrust investigations typically drag on for years — often so long that markets have already moved on by the time enforcement lands.
On 8 February, the Commission sent Meta a Statement of Objections and notified it of possible interim measures. This essentially warning it may order Meta to reverse the policy temporarily while the broader investigation continues. This came just two months after the Commission opened formal proceedings on 4 December 2025.
According to Teresa Ribera, the EU’s competition chief:
“AI markets are developing at rapid pace, so we also need to be swift in our action.”
Translation: Brussels is not waiting years while the competitive landscape gets locked in.
Meta’s defence reveals the real question
Meta isn’t conceding. It says the WhatsApp Business API was built for business-to-customer messaging — order confirmations and appointment reminders, for example — not as a distribution channel for standalone AI chatbots.
“There are many AI options and people can use them from app stores, operating systems, devices, websites, and industry partnerships,” a Meta spokesperson said, adding that “the Commission’s logic incorrectly assumes the WhatsApp Business API is a key distribution channel for these chatbots,” according to Reuters’ reporting on the EU’s move.
It’s not an unreasonable argument on its face. But it exposes the question every platform vendor will soon face: where’s the line between integrating your own AI and locking out everyone else’s?
That question matters for WhatsApp because — despite the “business” label — it functions like infrastructure for customer communication in many markets. And it matters even more because WhatsApp is not the only platform trying to become the default home for AI assistants.
This isn’t just about WhatsApp
Look across the communications stack and you’ll see the same pattern playing out everywhere: AI is becoming the interface, and platforms are incentivised to make their AI the default.
Microsoft has Copilot embedded throughout Teams. The company has also signalled support for third-party agents via Model Context Protocol (MCP), allowing agents to work with third-party apps and services in Teams environments. But Copilot remains the privileged experience, and the licensing model creates strong incentives to keep users inside the Microsoft AI ecosystem.
Zoom’s AI Companion is similarly baked into meetings, chat, and contact centre experiences. Integrations exist, but the default pathway is Zoom’s own AI — which is exactly where a platform vendor wants user attention, usage telemetry, and monetisation to flow.
Salesforce, meanwhile, is building an entire agent platform with Agentforce as the orchestration layer. Third-party AI doesn’t get equal billing — it operates within the platform’s framework or not at all.
None of these platforms have faced the same scrutiny Meta is now under. But the principle the EU is trying to establish — that dominant communications platforms can’t privilege their own AI by excluding rivals — doesn’t stop at WhatsApp. If interim measures are imposed and the precedent sticks, every vendor embedding proprietary AI into a dominant platform will need to answer the same question Meta is facing: how open is open enough?
As one legal analysis put it in the context of Meta’s WhatsApp integration, platform owners face a structural trade-off: “An open ecosystem maximises overall value but limits the dominant firm’s ability to capture value, while a closed ecosystem maximises value capture but increases innovation costs,” as discussed on the Kluwer Competition Law Blog. That trade-off just became a lot more expensive to get wrong.
What to watch, and what to do
Meta will respond to Brussels within weeks. An oral hearing follows, then a decision on whether to impose interim measures. That decision will matter for every platform vendor watching to see where the line gets drawn.
Beyond this case, the EU’s Digital Markets Act revision is due in May 2026, and recent legal commentary suggests AI is firmly on the agenda.
A Quinn Emanuel round-up of EU competition developments notes stakeholder support for bringing AI-related services into the DMA’s scope and flags the growing intensity of AI-linked competition enforcement (see “Key EU Competition Law Developments: 2025 Overview and 2026 Predictions”).
If your strategy depends on a single platform’s API (whether that’s WhatsApp Business, Teams, Zoom, or any other communications layer) you’re now exposed to two distinct risks: the platform can change its terms, and regulators can force those terms to change back. Either outcome can break integrations you’ve already deployed.
The answer isn’t to avoid embedding AI into your UC stack. The productivity gains are real, and users expect it. But the Meta case strengthens the argument for platform diversity.
The EU’s decision on interim measures will arrive in weeks, not years. What follows will determine whether UC platforms can treat AI integration as a product decision or whether openness becomes a regulatory requirement. For anyone managing communications infrastructure, that’s not an abstract question — it’s a variable in your next platform evaluation.
Read more:
