Meta’s AI chatbot finally launches in Europe with limited features

Meta AI’s delayed European launch marks a significant milestone for the company’s AI strategy, though with notable limitations compared to its global rollout. After extensive regulatory negotiations, European users will finally gain access to Meta’s chatbot functionality across the company’s major platforms, but without the image-based features available elsewhere—highlighting the ongoing tension between AI innovation and Europe’s stringent privacy regulations.

The big picture: Meta is finally rolling out its AI chatbot to 41 European countries and 21 territories this week, following a prolonged delay caused by privacy concerns and regulatory hurdles.

• The AI assistant will be available across Instagram, WhatsApp, Facebook, and Messenger, but with significantly reduced functionality compared to other regions.
• European regulators had previously blocked the rollout after concerns about Meta training its AI models on user data dating back to 2007 without explicit consent.

Key limitations: European users will only have access to text-based chatbot features, with image generation and visual recognition capabilities entirely absent.

• Users can ask questions, brainstorm ideas, plan trips, and surface certain Instagram content, but cannot generate or edit images or ask questions about photos.
• The model serving European users isn’t trained on EU user data, a critical concession to address regional privacy concerns.

Why this matters: The restricted rollout critically impacts the flagship functionality of Ray-Ban Meta smart glasses, which can’t offer their headline feature in Europe.

• The glasses’ key selling point—allowing wearers to ask the AI about what they’re looking at—will be unavailable in European markets.
• This creates a two-tier product experience between European customers and those in other regions where the full functionality is available.

Behind the decision: Meta described the European launch as following “intensive engagement” with regulators, suggesting significant compromises were necessary to enter the market.

• European privacy legislation requires companies to allow users to opt out of data collection, but regulators argued Meta should instead use an opt-in model for AI training.
• The compromise solution appears to completely exclude European user data from the training process rather than implementing either opt-in or opt-out systems.