Public trust in advertising spokespeople has deteriorated considerably over the past several years and generative AI is only part of the explanation. A series of high-profile cases in which well-compensated public figures endorsed products and platforms that subsequently collapsed, or turned out to be outright schemes, taught audiences a lesson that no amount of subsequent celebrity casting has undone.
The lesson is straightforward: a recognizable face attached to a product indicates that someone paid for the association and, most likely, nothing more.
That erosion has created a structural problem for advertisers. The traditional mechanism for transferring credibility to a product, borrowing it from someone the audience already trusts, works considerably less well when audiences have learned to interpret the borrowing itself as a warning sign.
Shane Ginsberg, Street Poller Media Founder and CEO, has pointed to that skepticism directly when explaining why an anonymous stranger on a sidewalk frequently earns more credibility than a compensated public figure. People do not trust a well-paid celebrity endorsing a product the way they trust a random person trying something for the first time on camera, with no visible financial stake in whether their reaction is positive. The stranger has no reason to lie, and more importantly, the audience can see they have no reason to lie.
Ginsberg has argued that this trust gap has widened further as more advertising becomes AI-generated or AI-assisted. Audiences increasingly cannot determine what is real and the rational response to that uncertainty is to grow more suspicious of anything that appears too polished, since polish is now the cheapest quality to manufacture.
Ginsberg connects this dynamic to a broader cultural shift, one where prediction markets and real-time sentiment tracking have moved toward the mainstream. According to him, people increasingly want an unmediated read on what other people actually think rather than what a curated influencer or a brand claims they think. Street polling, in this context, functions less like conventional advertising and more like a live temperature check on public opinion, with brand messaging traveling alongside it rather than driving it.
There is also a practical dimension driving this shift. Platform compliance systems across Meta and TikTok have grown increasingly hostile to claims-heavy advertising, particularly in regulated categories, which means the formats that read as most obviously promotional are also the ones most likely to be rejected before reaching anyone. A stranger describing a personal reaction occupies different territory than a brand asserting an outcome, both in how audiences receive it and in how review systems classify it.
That convergence, between what audiences find credible and what platforms will actually distribute, is part of why the format has expanded into almost every market imaginable.
