Overview
- Banks including Citigroup and Bank of America have issued formal policies that forbid third‑party AI agents from logging into customer accounts or executing transactions, and JPMorgan’s leadership has signaled similar limits.
- Companies cite concrete security failures such as agents reaching live systems, exposed credentials, and anti‑fraud triggers that can let bots initiate unwanted transfers or transactions.
- Retailers and services from Amazon to Delta and Yelp are using technical blocks, licensing rules, and litigation to stop agents from shopping or booking on users’ behalf, with Amazon’s legal fight against Perplexity as a high‑profile example.
- Web infrastructure firms such as Cloudflare have rolled out anti‑bot features that let sites detect or block agentic crawlers, producing inconsistent access and frequent CAPTCHA or checkout failures for users.
- Analysts warn the economic stakes are large: estimates show up to about $1.6 trillion in low‑cost deposits could be at risk if agents reallocate cash, a dynamic that has pushed banks, regulators, and standards groups to press for auditable controls and clearer liability rules.