Shoppers Embrace AI Agents for Grocery Purchases

PYMNTS Intelligence published findings in its Agentic AI Report showing growing consumer openness to AI-managed purchases. Per PYMNTS Intelligence, more than half of U.S. consumers already use AI during buying, nearly half said they would consider letting an AI agent shop for groceries, and 44% said they would trust an AI agent to select gifts. PYMNTS reports that in January Google launched an open-source standard to enable agentic commerce and that Google has begun integrating agentic commerce into its ecosystem with plans that include Shopify, Target, and Walmart. PYMNTS also reports that consumers express significant privacy and payments concerns around agentic commerce.

Agentic commerce combines recommendation, decision automation, credential delegation, and payments flows. Companies and platforms enabling autonomous shoppers will need to integrate identity verification, secure credential delegation or tokenization, and fine-grained consent mechanisms. Industry-pattern observations note that payment ecosystems evolve through standards and APIs to reduce direct credential sharing; practitioners building agentic flows typically rely on tokenized payments, scoped API keys, or delegated-authority primitives to limit exposure of primary payment instruments.

Industry context: Consumer acceptance is a necessary but not sufficient condition for widespread agentic commerce. Reporting by PYMNTS shows user intent leaning toward delegation, but simultaneous privacy and fraud concerns create a high bar for safe, usable implementations. Observed patterns in similar shifts suggest adoption often trails behind robust controls for auditability, revocation, and liability allocation across platforms, merchants, and payment processors.

Indicators that will matter to observers include adoption of standardized delegated-payment APIs from major payment networks, announcements of scoped-credential tooling from cloud or payments vendors, pilot programs from large retailers (Shopify, Target, Walmart as named by PYMNTS), and consumer-protection rules or guidance from regulators. For practitioners, watch for developer SDKs that implement consent-first flows, transaction-level visibility, and easy revocation of agent permissions.

PYMNTS Intelligence documents rising consumer willingness to hand routine purchases to AI agents while flagging payment and privacy frictions that could shape product design and infrastructure choices for merchants, payment providers, and platform developers.