Walmart recently announced it will allow artificial intelligence agents to perform shopping tasks on behalf of its customers through new "Agent Profiles." This strategic shift puts the retail giant at odds with Amazon, which continues to block third-party AI tools from accessing its marketplace.
Walmart's $200 spending limits for Agent Profiles
Within the next month, Walmart plans to introduce Agent Profiles, a feature that allows users to register AI assistants as distinct entities within their accounts. According to the report, this system is designed to separate automated purchases from human-led shopping, functioning similarly to how a primary credit card holder adds an authorized user to an account.
To mitigate the risks of autonomous spending, Walmart is implementing customizable financial guardrails. Customers can choose to require manual approval for every single transaction or set specific spending ceilings—such as $50, $100, or $200—that an AI agent cannot exceed without further permission.. This granular control is a key part of Walmart's attempt to make autonomous shopping feel safe for the average consumer.
The Perplexity lawsuit and Amazon's ban on Muse
While Walmart opens its doors, Amazon is aggressively securing its perimeter. as reported, Amazon has blocked Meta's Muse agent from its site, informing users that unauthorized AI agents violate the company's conditions of use. This restrictive stance is paired with a push for Amazon's own proprietary tools, including Alexa for Shopping and the "Buy for Me" feature.
The tension between the e-commerce leader and AI developers has already escalated into the courtroom. In November 2025, Amazon filed a lawsuit against the startup Perplexity, alleging that Perplexity's Comet browser agent was used to covertly execute purchases on the Amazon marketplace. By restricting agents from Google and OpenAI as well,Amazon is signaling that it views third-party AI agents as a threat to its data moat and transaction control.
From ChatGPT to Claude Duet and OpenAI's Dots
The current struggle between Walmart and Amazon is part of a broader industry evolution moving beyond the conversational chatbots popularized by the launch of ChatGPT in late 2022.. The inudstry is now pivoting toward "agents"—programs capable of executing multi-step workflows, such as comparing prices across different sites and completing a checkout process without human intervention .
Several tech titans are racing to dominate this space. Anthropic recently previewed Claude Duet, an agent designed to operate across a user's connected applications to handle online purchases. Meanwhile, OpenAI launched its own always-on agents known as Dots last month. Meta's Muse has already seen significant traction, with Sensor Tower data indicating more than 10 million downloads in the United States and Canada since its September launch.
How Walmart will verify the security of third-party agent access
Despite the convenience of Agent Profiles, several critical questions remain regarding the technical implementation of Walmart's open platform. It is currently unclear how Walmart will verify the identity of these agents or prevent "spoofing," where a malicious program might attempt to impersonate a registered AI assistant to bypass spending limits.
Furthermore, the source does not specify whether Walmart will charge a fee to AI developers for this access or if the company will collect a different tier of data from agent-led transactions. While Walmart's approach is more welcoming than Amazon's, the lack of detail on the API security framework leaves a gap in understanding how customer data will be protected when handed over to third-party agents like those from Meta or Anthropic.
Comments 0