Pinterest’s new AI tools for the discovery era: from ads to personalised shopping

As AI changes how people search, browse, and make decisions, the industry is moving beyond the traditional search-and-click model toward a more conversational and generative web. In this new environment, brands are competing not just for attention, but for recommendation, relevance and action.

Ahead of Cannes Lions, we are introducing new AI-powered tools built for this shift, including Business Assistant, Pinterest Model Context Protocol (MCP) and new Pinterest Performance+ creative capabilities.

We're also rolling out Ask Pinterest, a limited access app that will allow us to experiment with the next generation of AI-powered shopping experiences and learn quickly from our users.

"The future of discovery won't be driven by keywords alone. It will be shaped by context, taste, and trusted recommendations," said Lee Brown, Chief Business Officer, Pinterest. "Pinterest has a unique advantage because people come to our platform to plan, curate and take action on what they want to do next. We're building AI experiences and infrastructure that tap into those signals in more useful and relevant ways."

Business Assistant brings AI-powered guidance to advertisers

Pinterest is building AI into advertiser workflows with Business Assistant, an AI collaborator in Ads Manager and mobile, currently in a closed beta in the US. It combines a deep understanding of an advertiser's business with Pinterest's platform insights to help advertisers drive the best performance, building on the approach pioneered by Pinterest Assistant.

Like Pinterest, Business Assistant is visual. It doesn't answer with walls of text, but demonstrates breakout Trends in graphs and shows top Pins to promote. For example, if "clean beauty routine" searches are up 42% one week and an advertiser wants to join in, Business Assistant can represent how user interest grew, alongside real Pins leading the Trend to inspire their ads.

We've also brought Business Assistant to mobile, so advertisers get proactive notifications about Trends, performance status and optimization opportunities.

For illustrative purposes

Pinterest MCP: making Pinterest operable in AI-driven workflows

In addition to further integrating AI into the advertiser experience, Pinterest is also embedding AI into partner tools with Pinterest MCP (Model Context Protocol), an AI-native infrastructure that connects Pinterest to the copilots and agentic tools advertisers rely on. As AI transforms marketing, partners need a reliable, standardized integration.

Built on the Model Context Protocol, Pinterest's offering provides secure access to campaign, analytics and keyword insights. It grounds AI workflows in Pinterest's unique signals including taste, trends and intent, allowing partner copilots to provide platform-specific guidance directly within existing tools.

We are evolving Pinterest MCP alongside alpha partners like PMG, Pacvue, Dentsu, Havas, Innovid by Mediaocean and Omnicom's Jump450. Their feedback is shaping the protocol across reporting, analysis, campaign planning and execution, directly informing how we expand in the future.

"Pinterest MCP helps us integrate Pinterest directly into the workflows our teams are already building, making it easier to analyze performance, uncover insights and act on opportunities without switching tools," said Chris Ivey, President of Jump 450. "What makes Pinterest especially compelling is the strength of its intent signals. People come to Pinterest to plan what they want to do next, creating valuable context that helps marketers make smarter decisions."

Pinterest Performance+ creative: increasing creative variation through new AI models

As AI scales creative variation across platforms, brands need smarter selection models to turn more assets into stronger performance. Pinterest is introducing a new AI model to work with Pinterest Performance+ creative to deliver the right variation at the right moment.

Through dynamic creative selection, the model can evaluate a broader set of creatives and identify the variant most likely to perform for each ad impression. This moves performance optimization from the ad level to the asset level, delivering a more personalized experience for users and better results for advertisers. In testing, the new model increased click volume by 7.5% compared to the previous singular variant model.¹

In addition to a more powerful creative selection model, brands will also have access to new ad review tools and enhanced creative reporting breakouts, giving them greater confidence in how their creative appears and performs on Pinterest.

Ask Pinterest: AI experiments outside the Pinterest app

As consumer behavior around AI search and discovery evolves, we're creating a new space to rapidly test and learn. Ask Pinterest is a new experimental app for exploring more conversational, visual-first and agentic shopping experiences.

Ask Pinterest brings Pinterest's Taste Graph and the visual discovery experience beyond the core Pinterest app, using Pinterest's proprietary signals around taste, intent and preferences to power more personalized recommendations and inspiration.

Ask Pinterest is designed for more conversational, complex, multi-step decisions that don't fit neatly into a single search, like planning a dinner party on a budget, finding a gift that feels truly personal, or furnishing a room over time. The experience helps us explore how AI can better support those richer shopping experiences while retaining context across sessions.

What we learn from Ask Pinterest will help inform future AI-powered experiences across the main app.

Learn more at Cannes Lions

Together, these updates reflect how Pinterest is evolving its platform and infrastructure for a more conversations, personalized and AI-driven era of discovery.

Pinterest will be back at the Pinterest Manifestival at Carlton Beach Club at Cannes Lions 2026. RSVP to attend at PinterestCannes.com.