Walmart and Home Depot Set the First Rules for AI Shopping Agents

Walmart and Home Depot have published the first major retailer frameworks governing AI shopping agents — the software systems that browse, select, and purchase on behalf of consumers. Walmart will launch "Agent Profiles" in November with per-order spend caps of $50 to $200; Home Depot will require agents to disclose their identity and the customer they represent before completing any transaction. Together, these two frameworks are expected to become the industry baseline for agentic commerce entering 2027.

This is a practical reckoning with something that happened faster than most retailers planned for. AI shopping agents — software that customers authorize to buy groceries, replenish home supplies, or find the best price on a specific item — are already operating on major retail platforms. Walmart and Home Depot are the first to respond with formal policies rather than informal tolerance or outright blocks, according to PYMNTS.

What Walmart Is Building

Walmart's "Agent Profiles" give customers a structured way to authorize AI agents while defining exactly what those agents can do. The core mechanism is a spending cap: customers can set limits of $50 to $200 per order for AI agents acting on their behalf. An agent handling weekly grocery replenishment can operate freely within its authorized budget ceiling. An AI attempting to exceed that threshold requires customer confirmation before the transaction completes.

This design reflects a deliberate choice: Walmart is not blocking agentic commerce. It is formalizing it. By building "Agent Profiles" into its infrastructure, Walmart creates a customer-controlled permission layer that logs what AI agents did, within what limits, and on whose authorization. A retailer that builds a first-party agentic commerce framework gains visibility into every AI-driven transaction in a way that ad-hoc agentic access cannot provide — visibility that matters for fraud detection, dispute resolution, and customer relationship management.

The November launch puts the framework in place before the holiday shopping season, the first major commercial test of agentic retail purchasing at volume.

What Home Depot Is Requiring

Home Depot's framework takes a different approach: mandatory identity disclosure. Before any AI agent can complete a transaction on the Home Depot platform, it must identify itself as an AI agent and specify the customer it represents.

This addresses a trust problem that has emerged as AI agents have proliferated: retailers cannot currently distinguish a human customer from an authorized AI agent from an unauthorized bot without explicit disclosure requirements. Home Depot's policy creates a clear legal boundary. An agent that fails to disclose its identity — or misrepresents who it represents — is operating outside the retailer's terms of service, creating liability that falls on the agent operator rather than the retailer.

For contractors and property managers who purchase heavily from Home Depot and have begun using AI agents to manage purchasing across multiple job sites, the identity disclosure requirement creates an accountability chain. Every purchase is linked to a specific authorized agent representing a specific human customer — which matters when the transaction is later referenced against a job budget or disputed on a company account.

Why These Policies Matter for the Industry

Agentic commerce is not a distant hypothesis. AI shopping agents are deployed today by consumers using smart home platforms, by small businesses automating supply purchasing, and by property managers handling procurement across multiple properties. The volume is still small relative to direct human purchasing — but it is growing, and the holiday season that begins this month will be the first major commercial test of agentic purchasing at retail scale.

Without formal frameworks, retailers face a fragmented situation: AI agents operating under different assumptions, different authorization levels, and no consistent accountability structure. A retailer that blocks AI agents entirely loses convenience-driven customers. A retailer that allows them without rules has no mechanism to detect fraud, enforce spending limits, or resolve disputes.

Walmart and Home Depot are creating clarity, not just for their own platforms but for the industry conversation about what agentic commerce governance looks like. The two frameworks together address both consumer protection (spend caps) and retailer liability (identity disclosure) — the two most pressing policy questions in this space.

Other major retailers — Target, Best Buy, Costco — have not yet published comparable policies. They will face increasing pressure to do so as AI shopping agents become more capable and more widely adopted.

What the Policies Mean for Consumers

For consumers, the Walmart framework offers a practical safety layer. Spend caps mean AI agents cannot make large unintended purchases without confirmation. Agent Profiles create a persistent log of what an authorized AI did — useful for auditing AI purchasing behavior and disputing charges.

The Home Depot identity disclosure requirement may be less visible to individual consumers but matters for enterprise and contractor customers. It creates a formal record that an AI agent, not a human, made a purchasing decision — a distinction that has downstream implications for expense reporting, contract compliance, and any dispute that reaches arbitration.

Both policies acknowledge a reality that the broader retail industry has been slow to confront: consumers authorizing AI agents need explicit structural protections that did not exist when every purchasing decision was made by a human at a keyboard.

What to Watch

The critical test is November-December. If Walmart's Agent Profiles see meaningful consumer adoption — if customers actively set up authorized AI agents for holiday purchasing — the framework's design faces real-world stress testing for the first time at scale. Watch for edge cases: agents hitting caps unexpectedly, identity disclosure failures causing transaction blocks, or retailers without frameworks becoming preferred targets for unauthorized AI purchasing bots that exploit the absence of rules.

Home Depot's disclosure requirement will also set a legal precedent. Any retailer that adopts a similar policy creates a discoverable record that AI, not a human, completed a specific transaction — a distinction that will become significant when the first major agentic commerce fraud or dispute reaches litigation.