Uber Eats is rolling out a new AI feature called Cart Assistant that aims to turn grocery lists into ready-to-checkout baskets with far less tapping and searching. The tool, now in beta inside the Uber Eats app, is designed to shorten the jump from idea to purchase—exactly where most digital grocery trips bog down.
How the Cart Assistant Works to Build Smarter Carts
Shoppers start by choosing a participating grocery store in the Uber Eats app, then tapping a purple Cart Assistant icon on the store’s page. From there, they can paste a list or upload a photo—everything from a handwritten note to a screenshot of a recipe—and the assistant converts those inputs into a structured cart.
The system recognizes items and quantities, proposes reasonable size equivalents when none are specified, and suggests close matches when it encounters vague language like “breakfast cereal” or “mild cheese.” Users can fine-tune the basket by swapping brands, adjusting pack sizes, or adding complementary items before checkout.
Personalization is built in. Uber says the assistant leverages a customer’s previous orders to prioritize familiar products—your usual milk, the oat milk you tried last week, the snack brand you often reorder—reducing the time spent re-finding staples. It’s the digital equivalent of knowing exactly where your go-tos sit on a store shelf.
Consider a weeknight scenario: upload a screenshot of a pasta recipe, add a quick photo of a fridge sticky note, and the assistant compiles pasta, tomatoes, garlic, parm, plus the laundry detergent you routinely buy. A few taps later, the cart is ready with preferred brands and sizes.
Why It Matters for the Future of Online Grocery
Online grocery is a high-friction category. Unlike a single-restaurant order, baskets often span dozens of items, with substitutions and sizing choices multiplying the work. Industry research from the Baymard Institute regularly pegs average e-commerce cart abandonment near 70%, a sign that even small bursts of friction cost sales. Cutting the list-building step could meaningfully lift conversion and average order value.
On Uber’s side, the stakes are large. The company reported more than 150M Monthly Active Platform Consumers in 2023 across mobility and delivery, according to its investor materials. Even modest improvements in time-to-checkout or basket completion across that base translate into material volume for grocers and delivery partners.
For retailers, intelligent list-building can also help with out-of-stocks and substitutions. If a favored brand is unavailable, Cart Assistant can present pre-vetted alternatives in the same size or dietary profile, reducing last-minute changes by couriers and post-order refunds that frustrate shoppers.
The Competitive Context in Grocery and AI Ordering
Grocery has become a proving ground for conversational commerce. Instacart introduced Ask Instacart in 2023, powered by large language models, to answer food questions and suggest products. That same year, reports surfaced that DoorDash was testing DashAI to guide ordering. Both Uber Eats and DoorDash also launched ChatGPT integrations to streamline restaurant discovery and planning.
Uber’s Cart Assistant distinguishes itself by living inside a specific store’s catalog and taking structured actions on the cart itself. It accepts photos of handwritten lists or recipes, a practical twist that reflects how people actually plan meals and errands. The more modalities an assistant supports—text, image, and eventually voice—the less context a shopper has to rewrite for the app.
The move follows earlier AI investments on the merchant side, including AI-generated menu descriptions, enhanced food photography, and summarized reviews—signals that Uber wants AI touching both the supply and demand edges of marketplace friction. Bloomberg reported in 2023 that Uber was exploring an AI chatbot for ordering; Cart Assistant is a visible step in that direction for groceries.
Privacy, Accuracy, and the Cart Assistant’s Shopper Impact
Personalization depends on history, so data stewardship is in focus. Uber says it prioritizes items based on prior orders; consumers will expect clear controls to opt out, delete history, or limit personalization to specific stores—especially in regions governed by frameworks like GDPR or CCPA. Transparency around what is stored and how long it is retained will shape trust.
Accuracy is the other test. Interpreting messy handwriting and ingredient lists requires solid vision models and a robust product knowledge graph. The assistant must map “two cans crushed tomatoes” to the right size, brand, and pack count, then gracefully degrade when something is unavailable. Expect occasional mismatches early on, with user edits training better defaults over time.
Importantly, the assistant is additive, not a replacement for human pickers or store inventory systems. Shoppers still approve the cart before checkout, and couriers will continue to manage in-aisle substitutions. The goal is to reduce the cognitive load upfront, not remove the final human validation step that prevents surprise items from sneaking into orders.
What to Watch Next as Uber Tests Grocery AI Tools
Adoption will hinge on measurable gains: time saved from first tap to checkout, higher completion rates, and lower substitution-related refunds. Watch for Uber to add voice input, budget filters, and coupon stacking—features that early reports suggested were under consideration industry-wide—and to expand availability beyond the initial beta cohort.
If Cart Assistant proves it can reliably turn a scribbled list and a recipe screenshot into a precise grocery basket, it could become the default way many households restock. In a category where speed and accuracy earn loyalty, shaving even a few minutes could be the difference between an abandoned cart and a weekly reorder.