For more than two decades, the commercial logic of the web was shaped by a simple user journey: type a query, scan a list of links, click through to a destination website, and complete an action. Domains sat at the heart of this model. A domain name was the memorable entry point to a web property and, for many organizations, the primary gateway to customers, reputation, and revenue.

AI-powered search is rewriting that journey. Google’s AI Overviews place an AI-generated answer at the top of many results pages, with citations presented as supporting evidence. In parallel, conversational “answer engines” such as ChatGPT with search, Perplexity, and Google’s AI Mode are normalizing the idea that users can get what they need without visiting the underlying sources. Google has positioned this as a scaled shift: in Q1 2025, CEO Sundar Pichai said AI Overviews had “over 1.5 billion users per month,” highlighting strong adoption and momentum.

This shift raises a question for anyone working in the Internet ecosystem: if the click becomes optional, does the domain become optional too?

The click becomes optional, but the domain does not

Domains themselves are not becoming obsolete; they are being repurposed. In the AI era, the domain is moving from “the place users go” to “the identity and origin that machines—and increasingly humans—use to decide what to trust, cite, and transact with.” The new role of the domain is less about being the final destination and more about being the anchor that makes AI-mediated discovery, attribution, security, and automation possible.

Treat your primary domain as a long-term identity asset, not just as an address. In an AI-mediated web, your domain is the label attached to your reputation—whether or not a click happens.

The new AI search interface changes what “visibility” means

Google has been explicit that AI Overviews are no longer an experiment: they are a mainstream interface layered above classic results. But the prevalence of AI Overviews varies by dataset, geography, and measurement method. Ahrefs, analyzing 55.8 million AI Overviews across 590 million searches, reported that AI Overviews accounted for 12.8% of searches by volume in its index as of June 15, 2025. Pew Research Center, using observed behavior from March 2025, found AI summaries appeared in 18% of Google searches in its dataset.

What matters for the domain industry is not the exact percentage, but the direction: the AI layer is already substantial, and it sits above the classic list of links—often answering the question first. That changes what “winning” looks like. In classic search, visibility was correlated with SERP position and click-through rate. In AI-first search, “visibility” often means being cited (or paraphrased) inside the AI answer, not necessarily visited.

Pew Research Center illustrates the behavioral impact clearly. When users encountered an AI summary, they clicked on a traditional search result link in 8% of visits versus 15% when no AI summary appeared. Clicks on links inside the AI summary occurred in only 1% of visits. In other words: AI Overviews can turn many web pages into background material used to construct an answer, but not rewarded with a visit. There is an important nuance, though. Google also argues that these AI features can help users explore more links for complex questions, and notes that when users do click from results pages with AI Overviews, those clicks can be “higher quality” (for example, spending more time on the site). The ecosystem is likely to see both effects at once: fewer total clicks, but potentially more qualified visits when they happen.

This is where the domain’s role begins to shift. If a “visit” is no longer the default unit of value, the domain becomes valuable as something else: a source label, a trust signal, and a machine-readable identity that can travel into AI-mediated interfaces. So, start measuring “visibility” as a mix of traffic and citations/mentions. If AI reduces clicks, your domain’s brand signal inside AI answers becomes part of your distribution strategy.

The citation economy: domains as the new imprint of authority

AI Overviews and conversational answer engines introduce a new kind of economy: a citation economy. In this environment, domains function like publisher imprints in academic literature. Even when users don’t click, the domain name (or the brand associated with it) is the marker of “who said this” and “where it came from.”

We can already see concentration effects at the domain level. Ahrefs reports that the top 50 domains account for 28.9% of all AI Overview mentions in its dataset. Pew similarly found that Wikipedia, YouTube, and Reddit were the most frequently cited sources in Google AI summaries and standard search results in its sample. Other studies, using different methodologies, surface somewhat different “top cited” lists—for example, Semrush reported Quora as the most commonly cited website in Google AI Overviews in its own AI search study. The specifics vary, but the pattern is consistent: AI interfaces tend to rely heavily on a relatively small set of domains.

This concentration matters because it suggests that, as AI interfaces scale, domain-level reputation and recognizability can become even more important than individual page-level performance. When an AI system chooses sources to ground an answer, it is evaluating a document, and it is evaluating the domain behind it as a publisher, curator, expert community, or reference hub.

At the same time, research suggests the citation economy is volatile. Semrush analyzed weekly citation patterns across more than 230,000 prompts over thirteen weeks (July 14 to Oct 12, 2025) across ChatGPT Search, Google AI Mode, and Perplexity, totaling over 100 million citations. Their results show that platform-level changes can rapidly reshape which domains are “winners.” In their dataset, ChatGPT’s citation frequency for Reddit fell from close to 60% of prompt responses in early August to around 10% by mid-September. Semrush also notes that Google AI Mode shows a clear preference for Google-owned domains, underscoring how platform incentives can affect attribution patterns.

The lesson for Internet professionals is that domains are not only competing for rank; they are competing to be included in—and resilient within—a platform’s sourcing and attribution logic. That logic can change quickly as products and retrieval methods evolve.

Domains as trust infrastructure in an AI-generated web

The domain industry can expand its narrative beyond “find a good name.” In the AI era, domains increasingly represent digital trust boundaries: they support verification, secure communication, and accountability. A memorable domain helps humans, and a well-governed, well-secured domain helps machines.

As AI makes it easier to generate convincing text, images, and video, it becomes harder to tell what’s real and who’s accountable. Gartner has argued that as generative AI reduces the cost of producing content, quality and authenticity will become focal points, with a greater emphasis on watermarking and other mechanisms to authenticate high-value content, partly driven by new regulation.

Domains help because they are a stable identity layer: they can be owned, controlled, and secured through DNS and digital certificates, which helps distinguish legitimate sources from lookalikes—something AI systems increasingly need when selecting sources and minimizing errors. And a domain is not just a URL: it also anchors email and APIs, supporting protections like SPF/DMARC and enabling safer machine-to-machine interactions as automation grows.

This is also where the operational side of the domain ecosystem matters. The impact of AI in the domain industry is tangible: AI is already being used to strengthen domain security and threat detection—identifying patterns and anomalies associated with phishing, DNS abuse, and other malicious activity. As AI agents begin to execute tasks (purchases, bookings, customer-service actions), domain-level trust and authentication become the substrate that determines whether a transaction is legitimate. Hardening your domain stack is now part of AI readiness. Prioritize domain security because AI will amplify both trust and impersonation risks.

Build a domain-first playbook for AI search

If we compress all of this into a single idea, it is that the domain is evolving from a “traffic destination” into a trust and coordination primitive.

In the classic web, a domain mainly needed to be memorable and routable. In the AI web, it must also be citable (a recognizable provenance label) and verifiable (a secure identity boundary). It also becomes a control surface. Google now explicitly frames AI Overviews and AI Mode as integrated parts of Search and points site owners to existing web controls—such as robots.txt and preview controls like nosnippet, max-snippet, and noindex—to manage how content is accessed and shown in search, including AI features.

AI systems also change how discovery works under the hood. Google explains that AI Overviews and AI Mode may use a “query fan-out” technique—issuing multiple related searches across subtopics—to assemble an answer. For domain owners, this reinforces a practical point: your domain is increasingly evaluated as a coherent knowledge base, not a collection of isolated pages.

For professionals in the Internet ecosystem, the implication is strategic. The question is no longer only “How do we get users to click our domain?”, it is also “How do we make our domain the canonical, trusted source that AI systems choose to cite—and that agents choose to interact with—while preserving an independent relationship with users?”

Domains are not disappearing in the AI era. They are becoming the identity layer that makes an AI-powered Internet governable, trustworthy, and commercially actionable.