Two assistants, one week
Instacart and Shipt each introduced AI-powered shopping assistants within days of each other, turning what had been a scattered set of retailer pilots into a direct feature-for-feature race between grocery delivery's two largest platforms. Instacart's version, called Clementine, lets a shopper describe what they want in conversation or upload a photo of what is already in the fridge, then builds a cart that accounts for real-time store inventory, dietary preferences such as gluten-free or organic, and prior purchase history. Instacart CEO Chris Rogers framed the pitch around a question every household asks nightly: what's for dinner.
Shipt's answer, Ask Shipt, covers close to the same ground: it identifies ingredients from a photo, plans meals against a stated budget, surfaces relevant promotions, and hands the shopper an editable cart before checkout rather than an auto-purchased one. Shipt Chief Technology Officer Smrutha Ipparthi described the goal as building practical innovations that fit naturally into everyday life, language that signals Shipt is positioning the assistant as a convenience layer rather than a wholesale replacement for how people shop.
Why both landed at once
The near-simultaneous launch is not a coincidence so much as a signal that conversational, photo-based cart building has crossed from experimental feature to table stakes for any grocery delivery platform competing for the same retail partners and the same subscription-fatigued households. Both companies are effectively betting that lowering the friction of list-making, the most tedious part of grocery shopping, is where AI investment pays off fastest, ahead of flashier but lower-frequency use cases like in-store navigation or checkout automation.
It also reflects competitive pressure neither company can ignore: Amazon, Walmart, and a growing set of retailer-direct AI assistants have all been pushing into the same conversational-commerce territory over the past year. Instacart and Shipt both depend on retailer partnerships for their core business, and an assistant that keeps shoppers inside their respective apps, rather than drifting to a retailer's own AI tool or a general-purpose chatbot, protects that intermediary position.
The retailer rollout matters more than the demo
Instacart's underlying Cart Assistant technology, which powers Clementine, is already live through retailer partners including Food Bazaar, Heritage Grocers Group, and Woodman's Markets, with Aldi U.S., Save Mart Companies, and Stew Leonard's confirmed to launch soon. That distribution list matters more than the consumer-facing demo: it shows Instacart is embedding the assistant directly into regional grocers' own storefronts and apps, not just its own marketplace, which extends the technology's reach well beyond Instacart-branded shopping.
Shipt has kept its retailer rollout list for Ask Shipt private so far. Instead, the company is layering on Shared Lists, a feature that lets multiple household members contribute items to one collaborative list that converts directly into a cart. That is a distinct wedge from Instacart's inventory-aware, photo-based approach: Shipt is betting on household collaboration as its retention hook, while Instacart is betting on inventory accuracy and personalization depth to keep shoppers coming back to Clementine specifically rather than a rival assistant.
What this means for grocery CIOs
For a grocery or CPG-adjacent retailer evaluating its own AI roadmap, the practical question is no longer whether to add a conversational shopping layer, it is whether to build proprietary technology, license a platform like Instacart's Cart Assistant, or wait for Shipt-style capabilities to reach parity through a delivery partner already integrated into the stack. Each path carries a different data ownership and customer-relationship tradeoff: licensing gets a retailer to market fastest but routes shopper behavior data through a third party's model.
Inventory accuracy is the hidden dependency behind both launches. Clementine's pitch depends on knowing real-time store-level stock, and any retailer partnering into Instacart's assistant needs inventory feeds accurate enough that an AI-recommended cart does not fail at pickup or delivery. That requirement pushes inventory management modernization higher on the priority list for any grocer weighing whether to plug into either platform, since a shaky feed undermines the exact convenience the assistant is supposed to deliver.
The competitive stakes beyond grocery
Both launches sit inside a broader agentic commerce push that has pulled in OpenAI, Google, and Amazon over the past several months, each racing to own the moment where a shopper hands a purchase decision to an AI system rather than browsing a storefront themselves. Grocery is a useful proving ground for that broader shift because purchase frequency is high, basket composition is predictable, and the value of getting recommendations right compounds weekly rather than during occasional big-ticket purchases.
That makes Clementine and Ask Shipt more than feature launches. They are data-gathering exercises that will shape how both companies' underlying AI models perform against each other over the next several quarters, and the retailer partnerships each platform locks in now will be harder to dislodge later, once shoppers habituate to one assistant's particular cart-building style over the other's and switching starts to feel like relearning a household routine rather than picking a neutral tool.
What to watch
The near-term signal worth tracking is whether Instacart's retailer distribution list keeps growing faster than Shipt can match it with its own partner rollout, since retailer breadth, not consumer feature parity, is likely to decide which assistant becomes the default for a given household. A second signal is whether either company publishes usage or conversion data tying the assistant directly to basket size or order frequency, which would give CIOs outside grocery a concrete benchmark for justifying similar AI investment in their own conversational commerce plans.
Retailers currently sitting on the sidelines of this specific arms race should treat this week's dual launch as the moment the conversational shopping assistant category matured past pilot status and into direct competition between well-funded platforms. Their category almost certainly needs one at this point. The decision in front of them now is which partner's approach, inventory-aware personalization along Instacart's lines or household collaboration along Shipt's, fits their own customer base and existing operational data better.
