AI search trust is already showing cracks. The timing could not be worse for marketers who have spent the past year optimising for it. ChatGPT, Perplexity, and Google AI Overviews are pulling meaningful share of product-research queries away from traditional search. Moreover, the trend line is moving in a direction the major platforms cannot ignore. The uncomfortable finding from the latest consumer research is not that people are using AI search — they are. Instead, the problem is what happens to consumer trust the moment they learn the answer was generated by AI.
As a result, this shift is changing how brands need to think about content delivery. As AI agents become the customers themselves, the provenance of content — who made it, how, and whether brands disclosed it — is becoming a competitive variable, not just a compliance checkbox.
The Disclosure Problem Marketers Are Not Ready For
Trust drops sharply once consumers see an AI label on content. Specifically, the same content that performs well under an editorial or brand-produced label underperforms by meaningful margins when the AI origin is visible. Furthermore, a companion finding makes the picture worse. Consumers actually prefer AI-generated content when it carries no label at all. The implication is straightforward and unflattering — AI content performance depends, in measurable part, on consumers not knowing its origin.
Advertisement
300 × 250
Notably, this is not only a consumer behaviour problem. Buyers are already tightening filters against AI-generated prospecting in sales contexts. Consequently, the same scepticism is now reaching marketing content at scale.
For marketers running generative engine optimisation programs in 2026, the AI search trust problem is now an active regulatory question — not a future consideration. The FTC and the European Commission are both pushing toward more aggressive labelling requirements for AI-generated commercial content. Similarly, state-level activity in California and New York is moving in the same direction. Critically, the EU AI Act is becoming the default global standard for AI disclosure — even for US organisations that believe they operate outside its jurisdiction. Therefore, the strategy that works today has a finite half-life before disclosure rules catch up.
The Three-Part Playbook for AI Search in 2026
The honest playbook for the next twelve months has three components.
First, build GEO and AEO capabilities. Consumer research on those surfaces will keep growing regardless of trust dynamics. Additionally, understanding how AI agents are reshaping marketing operations is now a baseline requirement for any team running content at scale.
Second, separate brand-owned editorial work from AI-generated optimisation work. Teams must keep this separation clean enough to comply with disclosure rules without dismantling the brand. For instance, Adobe’s all-in bet on agent-based marketing already builds this separation into platform architecture. Buyers should demand the same from every vendor in their stack.
Third, invest in source-attribution and provenance signalling now. Platforms that surface AI-generated answers will eventually surface provenance metadata too. Consequently, brands that own that metadata will outperform brands that do not. This connects directly to the privacy-first identity and source-attribution infrastructure that forward-looking teams are already building. In addition, AI agents present unresolved identity and governance challenges that extend into content provenance. Teams that solve governance solve attribution at the same time.
In summary, marketers who treat AI search as a pure traffic-acquisition channel are heading toward the wrong side of both regulatory and consumer-trust shifts. By contrast, marketers who treat it as a trust-building channel that uses AI as a delivery mechanism are positioning for the version of the channel that survives the disclosure transition.
Reporting based on consumer research summarised by Constantine von Hoffman at MarTech.org.
