
Meta Joins Retail and Payment Firms to Set Rules for Shopping Bots
A new open protocol aims to verify autonomous artificial intelligence buyers before transactions cause widespread fraud and merchant pushback.
7 Oct 2026
A coalition of tech and retail companies has agreed to establish common operational guidelines for automated purchasing programs. Meta Platforms has joined Walmart, the payments processor Stripe, and enterprise startup Sierra Technologies to define how automated software assistants should transact with merchant websites.
The effort centers on a framework dubbed the Personal Agent Protocol. Created to function across different platforms, the code provides a verification handshake that lets online storefronts confirm whether an automated visitor genuinely represents a specific paying human customer.
The push arrives as technology developers promote agentic commerce, where customers delegate product searches, price comparisons, and checkout steps entirely to software assistants. Handing software direct access to bank accounts and payment cards, however, has outpaced existing merchant security measures.
Banking institutions have grown increasingly wary of the trend. A group of six large commercial lenders, including Capital One and Bank of America, recently published a joint appeal demanding strict oversight, warning that unregulated automated buying could trigger surges in fraudulent transactions and billing disputes.
Financial institutions also noted that software buyers might quietly favor particular brands or payment rails based on merchant commissions rather than customer preferences. The banks argued that common rules must guarantee user choice, data protection, and transparency across competing platforms.

Backers of the protocol liken the technical challenge to the early days of electronic mail, which succeeded because independent servers agreed on a single communication format. Without clear standards, retail servers struggle to tell legitimate digital assistants apart from malicious web scrapers.
That friction has already caused fractures across the retail sector. Amazon blocked Meta's personal digital assistant, Muse, from operating across its marketplace after the tool failed to identify itself while trying to interact with customer account records.
Meta has faced technical fallout on other fronts as well. Security researchers disclosed multiple software flaws that engineers hurried to mend immediately before the assistant's wide release, while subsequent reports highlighted how the software gathered background details on individuals who had never registered to use it.
Financial stakes remain high for social networks hoping to control the primary doorway to electronic retail. Financial analysts at Citigroup estimate that Meta's shopping assistant could generate upward of twenty-seven billion dollars in revenue before the decade closes.
Sierra co-founder Bret Taylor, who also chairs OpenAI, noted that shopping platforms face disorder if retailers cannot track what autonomous code does on their sites. Backers now hope other major artificial intelligence creators will accept the protocol to prevent merchants from blocking automated buyers entirely.