Home Depot Introduces New Policy for Artificial Intelligence Agents Accessing Its Services

Home Depot is set to introduce a new policy for artificial intelligence (AI) agents accessing its services. The move is aimed at ensuring that these AI-powered tools are legitimate and acting on behalf of their human owners.

According to Angie Brown, executive vice president and chief information officer at Home Depot, the company wants to verify that an AI agent is not just a bot trying to mimic human behavior. Instead, it needs to confirm that the agent has a genuine purpose and is authorized by its owner to interact with the retailer's website or apps.

Home Depot's decision comes as several tech companies have launched consumer-focused AI agents in recent times. These agents, such as Meta's Muse and Spear Street Technology's Instinct AI, promise to simplify online shopping by scanning products, comparing prices, filling carts, and making purchases on behalf of their owners with minimal human intervention.

The rollout of these AI agents has been met with mixed reactions from online retailers, with some welcoming the technology while others express concerns about its potential misuse.

Home Depot has expressed reservations about allowing AI agents like Muse onto its website without proper identification.

The company's concerns are not unfounded, given that Amazon recently barred Muse from shopping on its marketplace due to the bots' failure to properly identify themselves.

Meanwhile, other retailers such as Walmart and Best Buy have announced integrations or openness to AI agents, while Meta has co-developed a personal agent protocol with Sierra AI to standardize AI identification.

Home Depot's reluctance stems in part from its loyalty program, Pro Rewards, which caters to frequent shoppers like contractors and remodelers who are eligible for discounts.

Home Depot's cybersecurity team breathed a sigh of relief after the initial warning period passed without major incident on its main website. However, the experience was still chaotic for those involved in responding to it.

The company has since joined Anthropic's Project Glasswing initiative, which provides selected companies with access to Claude Mythos Preview to help secure critical software. Home Depot is one of a few retailers to have been given this opportunity, as most slots were reserved for banks and cybersecurity firms.

Home Depot's participation in the project allowed its AI agents to scan its code and identify vulnerabilities that needed fixing. While the review did uncover some genuine issues, it also flagged a few false positives, requiring the team to revisit and correct these errors.