Trust in AI search is slipping, users are cross-checking answers on more sites, and AI visibility is now more closely linked to brand authority than to classic SEO metrics. These are some of the main takeaways from new research by Fractl and Search Engine Land, shared by Fractl cofounder Kelsey Libert at SMX Advanced. The report provides an in-depth view of how consumers judge AI-generated responses, which signals shape AI recommendations, and where brands are lagging on governance and transparency. After the event, I spoke with Libert to unpack the findings further. We discussed the widening trust gap in AI search, the importance of earned media and entity authority for AI visibility, and why many companies remain ill-equipped for the operational demands that AI brings. The honeymoon is over Libert’s top-line finding is difficult to overlook. In 2025, 82% of consumers said AI search was more useful than traditional search results. By 2026, that number had fallen to 54%—a 28-point drop in just one year. Over the same period, the share of skeptics increased sixfold. I asked Libert what she believes is behind this decline. “Hallucinations. At first, AI felt like a seamless instant-answer engine that outperformed Google’s cluttered SERPs. As people started doubting AI responses and had to spend more time verifying them, that instant gratification vanished, and the perceived helpfulness plummeted,” Libert explained. Still, she isn’t completely negative about the future. “AI is on an exponential improvement…