What it takes to be trusted in the age of AI Search
author: M. Slivinskiy, 22 August, 2026
Search hasn't changed its underlying logic, but it has got harder to satisfy.
For years, brands treated search visibility as a technical problem: fix the crawl errors, optimise the metadata, hit the right keyword density, and traffic would follow. That kind of optimisation is now a baseline, not a strategy.
Generative search has changed what visibility looks like, even though the best, most relevant, highest-quality content still wins. The difference is that AI-powered search engines are now trying to find the single best answer among many relevant options. Here's what it actually takes to be trusted in that environment.
Trust is the new ranking factor
When a generative system builds an answer, it isn't just matching a page to a single query. It selects the sources it considers reliable, synthesises the most relevant and substantive material, and presents a conclusion to the user. The question every brand, publisher, and content team needs to ask is simple: are you the kind of source that gets selected?
How AI search engines evaluate trustworthiness
Trustworthiness isn't really a single visible signal. It is better understood as a pattern of quality across several dimensions. The EPOS framework, employed by Yandex Search, is a useful way to think through that pattern: Expertise, Practicality, Originality, and Substance.
Expertise
Content written by or clearly attributed to a real human expert reads differently than generic material produced at volume. Expertise needs to be demonstrable, not just claimed in an author bio. A named author with verifiable, relevant credentials is one of the clearest ways to make that expertise visible.
Practicality
A page that answers the literal query, but leaves the underlying decision unresolved, is worth less than one that anticipates what a user will actually need next. Good content should address the stated query directly, but more importantly, it should help someone actually finish the task they came for.
Originality
Content that repeats what dozens of other pages already say gives a generative system little reason to select it. A distinct angle, a genuine point of view, or an original way of resolving a common problem all give a source something to be chosen for, rather than being one interchangeable option among many covering the same ground.
Substance
Substance is about whether content has real accuracy and depth, the kind that holds up under scrutiny. A generative system has more reason to trust a source that is substantive enough to resolve the user's problem than one that reads well but doesn't hold up under close examination.
The information footprint
Since a generative answer draws on several sources at once, trust-building requires a multi-channel approach. Brands need to think about their full information footprint, meaning the entire ecosystem of places where they're described, reviewed, compared, and discussed. This includes independent reviews, expert comparisons, forum discussions, blogger content, niche publications, and more. Broadening that footprint means real PR strategy, a genuine presence in expert communities, and relationships with credible third-party voices.
The content quality bar has moved
Generative search does not eliminate the underlying logic of content marketing. It raises the bar for what "solving the task" means. Search has always rewarded content that resolves a user's need; what's different now is how broadly that need gets interpreted. A generative system, such as Alice AI that is embedded in Yandex Search, weighs not just the stated query, but a user's likely intent and the adjacent questions they haven't yet asked. Content that answers one question adequately is worth less than content that resolves the broader need behind it.
Focus beats volume
Under competitive pressure, it's tempting to just publish more. That logic made sense in an earlier, less saturated era of search, but the better strategy now is focus over volume. Identify the audience whose problem your brand is genuinely best positioned to solve, and make the case for why. Generative search rewards completeness, clarity, and usefulness — those are the qualities worth building toward, not sheer output.
Search visibility is everyone’s job now
One of the biggest shifts required to earn trust in generative search is organisational. Search performance is now shaped by product decisions, the actual strength of the value proposition, the credibility of the people behind the content, and the trust signals built through PR, partnerships, and community presence. Visibility can no longer be treated as the SEO team's job alone. Trust has to be earned across the whole business.
Conclusion
In the era of SEO, the brands that earn a place in generative answers are the ones that consistently demonstrate real expertise, usefulness, original perspective, and depth: the same qualities EPOS was built to capture. None of that comes from a single team or a single page. It's built by everyone shaping how a brand actually shows up on the web, and it's that combined effort, not any one optimisation, that earns trust in the end.
Author Bio
Mikhail Slivinskiy is Search Ambassador at Yandex with over 15 years of experience in search technology and SEO. At Yandex, he has worked across product development, webmaster tools, and publisher engagement, including leading Yandex Webmaster from 2017 to 2024. He now focuses on how AI-driven search is evolving and how businesses can maintain visibility through authoritative content.