Charles Schwab, Kohl's and Neiman Marcus are among the companies moving to stop personal AI agents such as Meta's Muse from acting on their sites, alongside JPMorgan Chase and Amazon, CNBC reported Sunday. The spread from one bank to brokerage and retail shows that companies now need a deliberate answer to software that shops and manages money on a customer's behalf.
What Happened
Personal agents are AI assistants that carry out tasks for a user, such as hunting for better savings rates, buying goods or canceling subscriptions. JPMorgan Chief Executive Jamie Dimon has said the bank blocks outside agents from acting for retail customers because of security risks, as The Primary Wire reported in October, citing Fortune. Amazon told CNBC that third-party apps buying on customers' behalf should "operate openly" and respect companies' decisions, and said it has asked Meta to remove Amazon from Muse. CNBC carried no response from Meta.
Walmart is taking the opposite course and opening its site to agents, saying they see the same prices human shoppers do.
An audit of 51 large U.S. retailers by Billions Network and AmanTrue, cited by CNBC, found that only 11 had taken a deliberate position on agents and five, including Amazon, Kohl's and Neiman Marcus, explicitly restrict them. Another 37 block shopping agents mostly by accident, through older bot rules written to stop data scrapers. None offers a way for a verified agent to complete a purchase. The audit is vendor research, and it notes that published rules do not prove technical enforcement.
Why It Matters
When an agent acts inside a customer's account, it blurs who authenticated, who authorized a payment and who is liable if something goes wrong. Michael Jabbara, who leads payment ecosystem risk at Visa, told CNBC the fraud risks are "very real and very tangible," citing fake agents built to steal credentials and legitimate agents that act beyond what a customer intended. The key questions, he said, are "who the agent is, who the agent is acting on behalf of."
Blocking may not hold. Ed Felten, a former chief technologist at the Federal Trade Commission, told CNBC that clicks an agent makes in a user's own browser are hard for a remote site to tell apart from a person's. That makes agent identity standards a competitive question as well as a security one. Banks have already found unauthorized agents inside customer accounts, The Primary Wire reported, citing Reuters and Axios.
What to Watch
Watch whether Meta changes how Muse identifies itself to websites, whether more banks and retailers publish formal agent policies, and whether Congress acts on calls from Sens. Mark Warner and Bernie Sanders for legislation on bots after reports of agents going rogue in customer accounts.
