AI Search, GEO and SEO: What Is Really Changing in Search in 2026

Online search is changing at a pace that is increasingly difficult to keep up with. For years, SEO largely meant improving rankings on a search results page dominated by links. In 2026, that view is no longer enough. ChatGPT, Gemini, Perplexity and Google AI Mode are turning search into an experience where the answer may appear before the user visits a single website.

This evolution means brands need to think beyond traditional rankings. Continuing to appear prominently on Google remains essential, but it is becoming equally important to understand when a brand is mentioned, recommended or used as a source by artificial intelligence systems. This is where concepts such as AI Search, GEO - Generative Engine Optimization - and optimisation for large language models are becoming part of a broader search strategy.

Rather than attempting to follow every technological update, businesses need to identify which changes have genuine implications for visibility, reputation and customer acquisition.

The model generating the answer can also change brand visibility

In August 2026, Google introduced Gemini 3.7 Flash as an option within Google Search AI Mode for Google AI Pro and Ultra subscribers, initially in English. The model had been officially introduced by Google and positioned as an evolution focused particularly on instruction following, reasoning and agentic tasks.

At first glance, changing the underlying model may appear to be a purely technical detail. For anyone responsible for digital visibility, it is not. Different models may interpret the same query differently, select different sources and structure answers in different ways.

The practical implication is clear: AI search monitoring should not be treated as a one-off snapshot. Brands need to observe trends over time, compare answers and identify changes in mention frequency, cited sources and recommendation patterns.

AI-referred traffic is starting to demonstrate commercial value

The growing use of artificial intelligence platforms is no longer simply a behavioural trend. According to data published by Shopify for the second quarter of 2026, sessions referred by artificial intelligence tools increased by 197% year on year, while sessions from organic search grew by 12% from a considerably larger base.

The same analysis shows that organic search still generates more total traffic than all the AI platforms measured combined. The emerging picture is therefore more nuanced: SEO and AI Search can work as complementary channels.

For businesses, traffic from ChatGPT, Gemini, Perplexity and other platforms should increasingly be analysed separately. A robust data analysis and analytics strategy can help distinguish genuine commercial growth from temporary curiosity surrounding new tools.

Being cited by AI is not the same as being recommended

One of the most important questions in GEO is deciding what should actually be measured. Counting references to a domain can be useful, but it does not reveal on its own whether a brand is earning preference.

An analysis published by Semrush, based on 1,094 categories monitored in ChatGPT between January and June 2026, found that achieving authority around a topic requires appearing consistently across multiple related queries rather than gaining visibility from a single prompt.

A citation can create visibility, while a recommendation can influence a decision. Brands therefore need to focus on building topical authority rather than simply accumulating mentions.

Topical depth is becoming more important

Rather than publishing disconnected articles, brands can benefit from building groups of related pages that address different questions within the same area of expertise. Taken together, these resources help search engines and generative models understand the relationship between a brand and a specific topic area.

This is also why SEO and organic search optimisation remain central. A strong GEO strategy does not replace fundamentals such as information architecture, relevant content, structured data, crawlability, authority and technical quality.

Search is becoming an ecosystem, not a single channel

The main conclusion is not that SEO is disappearing. If anything, the discipline is expanding. Information discovery now takes place across traditional search engines, generative answers, social media, video, specialist platforms and artificial intelligence assistants.

In this environment, the advantage will belong to brands that can build topical authority, reliable data, useful content, a consistent digital presence and the ability to measure emerging touchpoints.

AI Search and GEO are still evolving rapidly. Models change, interfaces are updated and new integrations continue to appear. Yet the strategic principle remains relatively stable: the better a business can demonstrate who it is, what it knows and why it is a relevant source within a particular subject area, the more likely it is to remain discoverable.