People will use AI to find products but not to buy them yet. What closes that gap?

There's a real split forming: people happily use AI to find and compare products, but most still won't let an agent actually buy. PwC called that trust gap the next barrier to agentic commerce. The jump isn't technical, it's delegating judgment. One-click checkout is easy; handing an agent the decision of what to buy is a different thing. And the failure modes are messy, an agent misreads a return policy, gets fooled by text on a product page, orders twice. When that happens, who answers, the merchant or the AI provider? People won't delegate if a mistake becomes theirs to untangle alone. The pattern that keeps surfacing: people will delegate the work, not the decision. They'll hand over the searching and ordering, as long as they can see what the agent knows about them, set what it's allowed to do, and reach a human when it screws up. Where's your personal line? A repeat grocery order within a budget, sure, but booking an expensive trip? What would make you trust it with the expensive stuff?