New data from Worldpay projects $261 billion of online spending will be done by AI agents in the next 5 years (9% of purchases).
44% of Americans say they would let an AI assistant browse for them – rising to 59% among 18–34,” the survey revealed, with most interest concentrated in finding the lowest prices, identifying best brands, and completing checkout on everyday purchases under $50.
And be that as it may, agentic shopping remains contested.
Yesterday, we learned Amazon is suing Perplexity over the agentic shopping feature on its AI browser Comet. Amazon claims that Perplexity is accessing accounts under the guise of human activity and insists that agentic-powered shopping degrades Amazon’s carefully built shopping experience for users.
Interestingly, just days prior, Amazon’s own AI shopping assistant Rufus was estimated to generate an additional $10 billion in annualized sales (disclosed by CEO Andy Jassy during its third-quarter earnings call).
If the browsing is automated, the buying can’t be far behind. But still, someone has to answer for identity and liability. As agents inch from “recommend” to “purchase,” the question of trust remains: Who authorizes the clicks? Who carries the credentials? And who audits the charge?
Today, Shubham Sharma shares Google and Mastercard execs’ thoughts on the new rails enabling agents to shop on our behalf — and explores what has to happen before chat becomes checkout.
–The Editors
