A Mastercard survey finds a third of teenagers trust an AI product pick over a friend's, and Rithum finds most young shoppers buy on AI recommendations without checking them. Influence is moving from the group chat to the model, at the exact moment Australia thins the social channel.
The Take: A generation of Australian shoppers is starting to trust a machine's product pick over a friend's, and that quietly rewires where influence lives. The recommendation is moving from the group chat to the model. Brands that are invisible to the model are about to be invisible at the moment of choice.
The numbers: In a Mastercard survey reported by Retail Dive, 31% of teenagers said they trust an AI product recommendation over a friend's and 23% would trust it over their parents. 27% said they are likely to use a fully AI-run shopping assistant that picks and buys for them, against 16% of parents. 62% of teens already use AI to research products at least monthly. Parents are not far behind at 49%, but the size of the generational gap is the tell about where this is heading.
Who wins: Separate Rithum research found 64% of 18 to 27 year olds have bought on an AI recommendation without checking it against another source. The brand the model names wins the sale before a website is ever opened.
64% of 18 to 27 year olds have bought on an AI recommendation without verifying it against any other source
What breaks: The referral, the review and the word-of-mouth loop all assumed a human passed the tip along. When the model passes it along instead, your presence in its training and its live retrieval becomes the new shelf position. Australia has just removed under-16s from most social platforms, so the channel that used to carry peer recommendation to teenagers is thinner at the exact moment AI moves in to fill the gap. Peer influence has not vanished. It has changed shape, and the new shape reads structured product data instead of a mate's text message.
For Australian operators: Treat the model as a sales channel, not a novelty. Make sure your product data, specs, pricing and structured detail are clean and machine-readable, because that is what an assistant reads before it recommends anything. Feed the reviews and third-party signals the models lean on, since those are what an assistant cites as proof. The under-16 ban does not protect your younger market from influence. It just changes who is doing the influencing.