Pinterest Unveils 'Ask Pinterest,' an Experimental AI Shopping Experience

On Wednesday, Pinterest unveiled a new experimental application called Ask Pinterest, giving the company a dedicated space to explore a more conversational approach to shopping and product discovery that could eventually become part of the main Pinterest platform. Alongside the new app, Pinterest also announced several AI initiatives, including Pinterest Model Context Protocol (MCP) for advertisers and new AI-powered advertising tools.

The announcements come just ahead of the Cannes Lions advertising festival, where artificial intelligence has become one of the industry’s central topics as brands, marketers, and technology companies increasingly focus on AI-driven advertising solutions.

Ask Pinterest is a web-based application that helps the company use its proprietary Taste Graph—this internal system maps users to their interests, preferences, and aesthetic tastes. Pinterest said the experimental app will initially be available to a limited number of users.

The AI-powered experience expands Pinterest’s traditional visual discovery model by introducing a conversational interface that lets users ask questions in natural language and receive personalised shopping suggestions, recommendations, and inspiration.

The launch also reflects a broader shift in online discovery, as AI chatbots continue competing with traditional search engines for consumer attention. Companies, including Google, have already integrated AI into shopping experiences for product recommendations, price tracking, and checkout assistance. At the same time, ChatGPT, Meta, Shopify, and others have also explored AI-powered shopping capabilities.

Rather than licensing its recommendation technology to outside AI platforms, Pinterest has largely focused on leveraging its own proprietary data to train AI models and develop AI-powered consumer experiences.

By releasing Ask Pinterest as a standalone application, the company gains the flexibility to experiment with new AI capabilities without affecting the core Pinterest experience used by millions of people.

Pinterest says the application is particularly useful for handling more detailed or multi-step requests that extend beyond traditional keyword searches. For example, users could ask the AI to help organise a dinner party, plan room décor over several months, or receive personalised shopping guidance based on broader lifestyle goals. The company said the objective is to evaluate how AI can improve shopping while maintaining conversational context across multiple sessions.

When users sign into their Pinterest accounts, Ask Pinterest can also incorporate their saved Pins and Boards to deliver recommendations that better reflect their individual interests and preferences.

Pinterest believes insights gathered through the experimental application will eventually help shape future AI-powered features for its flagship platform.

Alongside the consumer-focused launch, Pinterest also introduced several AI tools designed for advertisers. The company announced a beta AI assistant in Ads Manager for U.S. advertisers, designed to simplify campaign management and optimisation.

Globally, Pinterest introduced a new AI model within its Performance+ advertising platform that automatically selects the creative most likely to perform best each time an advertisement is displayed. In addition, the newly announced Pinterest Model Context Protocol (MCP) provides a standardised infrastructure layer that enables advertisers to manage and monitor campaigns using third-party AI agents and automation tools.

Discussing the announcements, Pinterest Chief Business Officer Lee Brown said the future of online discovery will rely less on keywords and more on context, personal taste, and trusted recommendations. Brown added that Pinterest believes its deep understanding of user interests provides the company with a unique advantage as AI reshapes how people discover products and ideas online.

Pinterest is currently available in limited access on both desktop and mobile web browsers, with the early rollout aimed at gathering feedback and refining the experience before considering broader availability.