Meta is opening the door for competing AI assistants on WhatsApp in Europe, but it won’t be free. The company will allow general-purpose AI chatbots from other providers to run on WhatsApp through its Business API for a limited 12-month period, charging per message. The move is aimed at blunting regulatory pressure while preserving a monetization path inside one of the world’s most widely used messaging apps.
For AI developers, the shift offers coveted access to WhatsApp’s massive reach — more than 2 billion users globally — under terms that could materially shape unit economics for chat services on mobile. For regulators, it’s a partial concession that tests whether opening core distribution with a toll is meaningfully pro-competitive.
- What Meta Is Changing in WhatsApp’s AI Bot Access Policies
- How WhatsApp’s Per-Message Pricing for AI Bots Works
- Regulatory backdrop behind Meta’s WhatsApp AI access move
- Who Can Participate and What’s Excluded in Europe
- What it means for the AI chatbot market on WhatsApp
- The key metrics to watch as rival chatbots join WhatsApp
What Meta Is Changing in WhatsApp’s AI Bot Access Policies
Until now, WhatsApp’s rules effectively kept general-purpose AI chatbots out of the Business API, curbing competitors to Meta’s own assistant while leaving room for customer-service automations that rely on approved message templates. Under the new framework, third-party chatbots — think models like ChatGPT or Claude — can operate conversationally via the API across European markets for a year-long window.
Critically, the opening comes with a meter. Meta plans to charge providers a fee for every “non-template” message sent, a category that covers the back-and-forth exchanges typical of AI chats. That means an interaction with dozens of turns can quickly add up, a stark contrast to the flat or usage-based pricing many AI companies apply in their own apps.
How WhatsApp’s Per-Message Pricing for AI Bots Works
According to company materials shared with partners, rates will vary by country and fall roughly between €0.0490 and €0.1323 per non-template message. On that basis, a 20-message conversation could cost a provider around €0.98 to €2.65; stretch to 30 messages and it climbs to about €1.47 to nearly €4. For a high-volume assistant, those figures can compress margins fast, especially when combined with underlying model inference costs.
Expect providers to adapt. Some will cap turns, summarize mid-chat, or switch to more compact prompts to contain spend. Others may introduce per-session prices, nudge users to subscribe, or push more interactions into lower-cost channels. A few may treat WhatsApp as an acquisition surface — offer a taste, then encourage users to move to owned apps where economics are friendlier.
Regulatory backdrop behind Meta’s WhatsApp AI access move
The European Commission had signaled plans to deploy interim measures over Meta’s earlier restrictions, arguing that blocking rival assistants from WhatsApp could distort competition while Meta promotes its own AI inside the app. Competition authorities in Italy and Brazil also opened probes into the policy after Meta announced it last year.
By temporarily allowing rival bots with paid access, Meta is effectively offering a waiting room solution while the broader antitrust investigation proceeds. EU officials have said they are assessing whether these changes reduce the need for emergency remedies and how they affect the underlying case. Given WhatsApp’s designation as a gatekeeper service under the Digital Markets Act, questions about self-preferencing and fair access won’t vanish with a time-limited pilot.
Who Can Participate and What’s Excluded in Europe
The opening applies to general-purpose AI chatbots that users can message directly on WhatsApp via the Business API. It does not restrict businesses that already use AI for customer support through approved, templated messages — those use cases were never the target of the original ban. The change mirrors a narrower allowance Meta previously tested in Italy, now scaled to Europe.
Practically, developers will still need a WhatsApp Business account and compliance with the platform’s policies, including data processing requirements. Distribution will likely run through verified business numbers, not a public “bot store.” That design choice matters: it provides access without fully treating WhatsApp as an open platform for AI agents.
What it means for the AI chatbot market on WhatsApp
For consumers, this could bring familiar assistants into WhatsApp’s chat list, reducing friction to try competing AI experiences in a space where people already spend time. For startups, it creates a new growth channel — but one that demands careful pricing and session design. Established players with stronger monetization funnels may handle the toll better than smaller entrants.
Strategically, Meta preserves two levers: it avoids the immediate hammer of interim enforcement while setting economic terms that deter pure free-riding on its distribution. If uptake is strong despite the fees, Meta can argue the market is functioning. If adoption stalls, regulators may scrutinize whether the pricing structure itself hinders effective access.
The key metrics to watch as rival chatbots join WhatsApp
Key signals over the next months will include how many major AI providers integrate, whether per-message pricing shifts toward per-conversation bundles, the extent to which developers pass costs to users, and any changes in the Commission’s antitrust posture. Also watch whether this 12-month window becomes a template beyond Europe — or a one-off concession that ends as quickly as it began.