Uber’s delivery unit, which includes meals, convenience items and groceries, reported $25.4 billion in gross bookings for the fourth quarter, representing a 26 percent year-over-year increase. Grocery orders remain a smaller share of the overall volume, yet executives have emphasized the category’s potential to raise user frequency and strengthen customer loyalty by turning the Uber Eats app into a broader shopping destination.
Practical Application of AI
Company leaders describe the Cart Assistant as an example of how Uber intends to integrate generative AI technologies into its core consumer experience. By anchoring the tool to a specific task—building a grocery basket—the firm aims to remove friction from a process that can be time-consuming on mobile devices. Praveen Neppalli Naga, who oversees consumer technology at Uber, said the initiative stems from observing how users jot down lists and then search for the same items within the app. Transforming that workflow into an automated feature, he noted, could shorten order times and reduce errors.
The assistant builds on earlier AI features introduced by Uber in July, when the company began offering restaurants automatically generated menu descriptions and summaries of customer reviews. Internally, engineers trained proprietary models on catalog data and order histories, while drawing on advances in computer vision to interpret images of cursive or printed handwriting. The system then cross-references recognized items with each retailer’s real-time inventory feed to confirm whether a requested product is in stock.
Industry-Wide Experimentation
Retailers across the sector are increasingly experimenting with large language models and automated agents to simplify shopping and personalize recommendations. Instacart last year debuted a suite of generative AI tools for grocers, while Amazon and Walmart are testing conversational assistants that can answer product questions or curate baskets. A 2023 analysis by McKinsey & Company estimated that generative AI could create between $400 billion and $660 billion in additional value for the retail and consumer packaged goods industries annually, largely by improving content creation, customer service and supply-chain planning.
For delivery platforms, predictive capabilities and seamless user interfaces are emerging as key differentiators in a crowded market. By automating cart assembly, Uber hopes to boost average order values and completion rates, two metrics closely watched by investors. In addition, the feature may encourage first-time grocery customers to experiment with the service, as it removes the need to search thousands of stock-keeping units on a phone screen.
Rollout and Availability
Uber said the Cart Assistant will appear as an in-app prompt for eligible users over the coming weeks. At launch, the experience supports text in English, with image recognition optimized for handwritten or printed lists photographed under typical indoor lighting conditions. The company plans to refine the model based on early feedback and may extend support to additional languages and retail partners later this year.
Retail chains participating in the initial deployment retain control over pricing, promotions and substitutions, mirroring the existing Uber Eats grocery framework. Orders are fulfilled either through in-store staff or third-party pickers, depending on each retailer’s operational model. Delivery fees, membership discounts and service charges remain unchanged.
Imagem: Internet
Competitive Landscape
DoorDash, Uber’s largest U.S. rival in meal delivery, has also accelerated its move into groceries and convenience items. Both platforms introduced on-demand alcohol delivery, partnered with pharmacy chains for over-the-counter medications and expanded white-label logistics offerings for merchants seeking to outsource same-day shipping. Instacart, meanwhile, has sought to defend its leadership position by deepening relationships with supermarket operators, licensing its technology stack and pursuing advertising revenue.
While all three companies promote AI as a lever for efficiency and personalization, Uber’s imaging capability marks a distinct approach. Analysts note that turning a paper list into a digital basket could resonate with time-pressed families who still rely on traditional pen-and-paper planning. The feature may also serve customers with limited typing ability or visual impairments if combined with voice dictation in future updates.
Regulatory and Privacy Considerations
As with other AI-driven services, the Cart Assistant raises questions about data governance and user privacy. Uber processes uploaded photos on its servers, extracting text and item identifiers before discarding the original image, according to information shared by the company. The firm stated it complies with relevant data-protection regulations and does not share raw inputs with external parties. Any personalized recommendations draw on aggregated order history and stated dietary preferences stored within a rider’s profile.
Industry observers are watching how regulators may respond to the growing presence of generative AI in consumer applications. The Federal Trade Commission has signaled interest in ensuring transparency around automated decision-making and preventing unfair practices. Delivery platforms that leverage AI for dynamic pricing or product promotion could face scrutiny if algorithms produce discriminatory outcomes.
Next Steps
Uber indicated that the Cart Assistant represents the first of several AI-driven shopping tools in its product roadmap. Future iterations might suggest recipe-based bundles, coordinate group orders among multiple household members or integrate with Uber’s ride-hailing service to provide store pickup options. For now, the company is monitoring key performance indicators such as conversion rate, average basket size and customer satisfaction as it measures the feature’s impact.
Although the timeline for a full international rollout has yet to be announced, Uber typically scales new offerings to Canada, Australia and selected Latin American markets after completing initial U.S. testing. Adoption rates in those regions will likely influence the company’s longer-term grocery strategy and help determine whether AI-generated carts become a standard expectation among app-based shoppers.
Crédito da imagem: Michael Nagle / Bloomberg / Getty Images
