ChatGPT Is Becoming a Search Giant: Is Google Losing Its Grip on the Internet?

THE JORNAL | Technology & Artificial Intelligence | November 5, 2026

For more than two decades, finding information on the internet followed a familiar ritual.

You opened Google, typed a few words, scanned a page of blue links, opened several websites, compared their information, and eventually decided which answer seemed trustworthy. If you wanted to know why your electricity bill had increased, how to repair a leaking tap, or which laptop was worth buying, Google gave you a list of places to investigate.

Today, an increasingly large number of people are doing something different.

They open ChatGPT and ask a complete question. They explain their situation, add details, ask follow-up questions, and expect an answer that brings the relevant information together. Instead of navigating ten websites, they may receive a structured explanation in seconds.

The difference seems small. One interface presents links; another presents an answer. But beneath that difference is a major change in how information is discovered, distributed, and consumed.

In November 2026, OpenAI's consumer product leadership made a particularly striking claim: ChatGPT now handles approximately one-fifth as many search-like queries as Google. The same reporting put ChatGPT's weekly active audience at approximately 1.2 billion people.

These are remarkable figures, but they require careful interpretation. One-fifth of Google's reported query volume does not mean ChatGPT owns 20% of the search-engine market. Nor does 1.2 billion weekly users mean that 1.2 billion people use ChatGPT as their primary search engine.

Understanding what these numbers actually measure is essential to understanding what is happening.

The central question is no longer simply whether AI can answer questions. It is whether the internet is moving from a system built around finding information to one increasingly built around receiving synthesized answers.

Key Takeaways

• OpenAI's consumer product leadership reported in November 2026 that ChatGPT handles roughly 20% of Google's search-like query volume. This is a company-reported comparison, not an independently audited market-share measurement.
• ChatGPT's reported weekly active audience has reached approximately 1.2 billion, according to the same November reporting. Weekly users and search queries are different measurements.
• Google's conventional search position remains enormous. Statcounter reported approximately 91.05% worldwide search-engine market share for Google in October 2026.
• Google is not standing still: AI Overviews, AI Mode, and Gemini are bringing conversational AI into its existing search ecosystem.
• AI-generated answers can reduce the need to visit individual websites, creating new challenges for publishers, independent journalists, businesses, and advertisers.
• Traditional SEO is evolving towards a broader discipline that includes visibility in AI-generated answers, source citations, brand recognition, and trustworthy information.
• AI answers are useful starting points, not guaranteed facts. Important claims should still be checked against original sources.

1. The Numbers Behind the Search Revolution

Before examining the companies competing for attention, it is important to understand the numbers being discussed.

There are three different measurements that are often mixed together in headlines:

1. Search-engine market share: The proportion of measured search-engine activity attributed to a particular search engine.
2. Search-query volume: The number of search-like questions or queries processed by a service.
3. Website traffic share: The proportion of visits to websites that originate from a particular service.

These measurements describe different things. A company can process enormous numbers of questions without sending an equivalent number of visitors to external websites.

That distinction is at the heart of the ChatGPT-versus-Google debate.

1.1 What Does the 20% Figure Actually Mean?

On November 3, 2026, Alex Kantrowitz of Big Technology published an interview with Adam Fry, OpenAI's consumer product lead for ChatGPT.

Fry explained that the volume of search-like queries handled by ChatGPT had reached approximately one-fifth of Google's query volume. He also revealed that ChatGPT had grown from 1 billion to 1.2 billion weekly active users in approximately 60 days.

Source: Big Technology, November 3, 2026.

The distinction is important.

OpenAI's reported comparison concerns questions that people traditionally might have submitted to a search engine but now submit to ChatGPT. According to Fry's explanation reported by The Verge, OpenAI counts queries that trigger a web search, rather than ordinary requests such as simple definitions or basic how-to explanations.

This means that the figure is not simply a count of every message written inside ChatGPT.

However, OpenAI has not published a complete, independently auditable methodology explaining every detail of its calculation. In particular, the precise treatment of repeated questions, follow-up prompts and multi-step conversations remains important when comparing its numbers with Google's.

The safest interpretation is therefore:

OpenAI says ChatGPT is now processing a volume of search-like queries equivalent to roughly 20% of Google's reported search volume.

It is a substantial milestone, but it should not be presented as proof that ChatGPT has captured 20% of all search-engine users or 20% of website traffic.

1.2 What Does 1.2 Billion Weekly Users Tell Us?

The second figure is equally impressive.

According to the November 3 interview, ChatGPT increased its weekly active audience from approximately 1 billion to 1.2 billion in two months.

That is an increase of around 200 million weekly active users, or approximately 20% relative to the earlier reported audience.

OpenAI had already publicly stated on August 31, 2026, that ChatGPT served more than 1 billion weekly active users. The November figure represents a newer disclosure from its consumer product leadership.

But weekly active users are not the same as people searching for information.

Consider a person who opens ChatGPT every day to:

• Write professional emails.
• Translate documents.
• Generate images.
• Explain difficult lessons.
• Write computer code.
• Organise a business project.
• Research a product.
• Ask questions about current events.

That person counts as a weekly active user, even if only a small proportion of their interactions involve searching the live web.

Consequently, the 1.2 billion figure measures the scale of ChatGPT as a product, not the size of its search audience.

1.3 Google's Five-Trillion-Search Benchmark

To understand the comparison, we need Google's own figures.

On May 21, 2025, Google publicly stated that its search engine was handling more than 5 trillion searches annually. The figure was also discussed during Alphabet's financial communications in 2025.

Source: Google, May 21, 2025.

A simple calculation gives:

Measurement	Approximate volume
Google searches annually	More than 5 trillion
Google searches monthly, using that benchmark	More than 416 billion
Google searches daily, using that benchmark	More than 13.7 billion
One-fifth of the annual Google benchmark	More than 1 trillion

The final row is a mathematical illustration, not a separately disclosed ChatGPT search count.

It assumes that OpenAI's comparison uses a directly comparable Google denominator. The companies have not published sufficient methodological detail to establish that every counting rule is identical.

There is another important development.

On October 28, 2026, Google CEO Sundar Pichai said that the year-over-year growth rate of Google Search queries was the highest in five years. Alphabet also reported that its Search and other revenue increased 12% year over year in the third quarter.

Source: Alphabet, Q3 2026 earnings communication.

This changes the interpretation of the competition.

ChatGPT is gaining importance, but Google's search activity is not necessarily shrinking. Google says its own search usage is growing rapidly, partly because its AI features encourage people to ask more questions.

The market may be expanding even as users redistribute some of their attention between competing services.

2. Why Estimates of ChatGPT Search Share Do Not Always Agree

A particularly interesting development is that independent industry researchers have produced a substantially lower estimate than OpenAI's reported 20% figure.

On October 28, 2026, SEO research company seoClarity published an updated model estimating that ChatGPT's search-equivalent activity represented approximately 7.5% of Google's search-query volume.

Its model estimated approximately 1.24 billion search-equivalent ChatGPT messages per day. It also calculated approximately 952 million search-like tasks after accounting for conversational follow-ups.

Source: seoClarity, October 28, 2026.

Why is this different from OpenAI's 20%?

The answer lies partly in methodology.

seoClarity used OpenAI's Signals v3.0 behavioural information, covering July to September 2026, and classified approximately 27.5% of ChatGPT messages as search-equivalent activity.

Its calculation included:

• 20.1% of messages classified as seeking information.
• 29.4% classified as practical guidance.
• An assumption that 25% of practical-guidance messages should count as search-equivalent activity.

The researchers then combined their estimate with an estimated 4.5 billion daily ChatGPT messages and a model of Google's search volume.

Importantly, the model assumed 20% cumulative growth in Google's search volume from its earlier five-trillion annual benchmark. That growth assumption was made by seoClarity; it was not a new Google-reported search count.

The two figures should therefore be presented differently.

Figure	Source	Measurement
Approximately 20%	OpenAI executive disclosure, November 3, 2026	Company-reported search-like query comparison
Approximately 7.5%	seoClarity, October 28, 2026	Independently modelled search-equivalent activity
More than 5 trillion annually	Google, May 21, 2025	Company's disclosed annual search benchmark

Neither figure should automatically be treated as a universal, independently verified market-share statistic.

A further complication is that an AI conversation can contain several questions.

Imagine someone asking ChatGPT to plan a holiday:

1. What is the best time to visit Japan?
2. How expensive is Tokyo?
3. Which neighbourhoods are best for families?
4. How do train passes work?
5. Can you compare three hotels?

Is that one research task or five searches?

Both descriptions are reasonable, depending on what is being measured.

Traditional search engines also generate multiple queries during a single research journey, but the conversational structure of AI makes comparisons more complicated.

For that reason, the most useful future measurements will need to distinguish between individual prompts, external web searches, completed research tasks, unique users and website visits.

Until companies publish comparable definitions, precise rankings should be treated cautiously.

3. Is Google Actually Losing Its Search Market Share?

This is where the discussion becomes particularly interesting.

A headline saying that ChatGPT has reached 20% of Google's search volume can create the impression that Google has lost one-fifth of its search market.

That is not what the available measurements demonstrate.

3.1 What Statcounter Actually Measures

According to Statcounter Global Stats, Google's worldwide search-engine market share was approximately 91.05% in October 2026.

Bing was second at approximately 4.8%, followed by Yahoo at 1.18%, Yandex at 1.14%, DuckDuckGo at 0.64% and Baidu at 0.42%.

Source: Statcounter Global Stats, October 2026.

These figures are useful, but there is a methodological detail that matters enormously.

Statcounter explains that its search-engine statistics are based on search referrals: visits to websites that originate from search-engine results.

It does not directly count every search performed.

Its network tracks billions of page views across more than one million websites. If somebody searches Google and never clicks a website, that activity does not appear as a search-engine referral in its measurement.

Source: Statcounter methodology and frequently asked questions.

Therefore, the 91.05% figure describes Google's dominance in Statcounter's measured search-referral activity. It is not a direct measurement of all global search queries.

It also does not include ChatGPT as though it were another conventional search engine in the same measurement category.

The important conclusion is that Google continues to dominate conventional search-engine referrals, even while AI assistants are gaining a larger role in information discovery.

3.2 Search Volume and Website Traffic Tell Different Stories

Consider two different situations.

A person asks Google:

What are the best ways to improve sleep?

Google displays a page of results. The person opens three websites and reads their recommendations.

That interaction can produce multiple website visits from one search.

Now consider someone asking ChatGPT the same question.

ChatGPT produces a complete explanation, perhaps with several references. The person reads the response and closes the application.

The second interaction might count as search-like activity for an AI service, but it produces no external website visit.

This is why search volume and referral traffic can tell very different stories.

One of the clearest recent examples comes from an analysis published by Search Engine Land on October 21, 2026.

Marie Haynes examined 12 months of Google Analytics 4 data across 100 websites, covering more than 145 million sessions.

Her analysis found that ChatGPT generated sessions equivalent to approximately 0.62% of the volume generated by Google organic search. When other large language model referrals were included, the figure increased to approximately 0.73% of Google organic traffic.

Source: Search Engine Land, October 21, 2026.

That is a very different picture from the 20% search-volume comparison.

It does not mean AI is unimportant. It means AI services currently send a much smaller volume of measurable visitors to these sampled websites than Google organic search.

The same study also found that, among 55 websites tracking conversions, ChatGPT visitors converted at 13.8%, compared with 3.02% for Google organic visitors.

Despite this higher conversion rate, Google organic search generated 27 times as many conversions as all the measured AI platforms combined.

These results are interesting, but they should not be generalised to every industry. The sample covered websites of varying sizes, and the study did not disclose every detail of its industry composition or conversion definitions.

3.3 AI Referrals Are Growing, Even From a Smaller Base

There is also evidence that ChatGPT is becoming a more important gateway to the wider internet.

A Semrush analysis published on April 7, 2026, examined more than one billion lines of US clickstream data from October 2024 to February 2026.

It found that outbound referral traffic from ChatGPT to other websites increased by 206% between January 2025 and January 2026.

The research also found that ChatGPT's web referral activity was distributed across a growing number of domains. More than 30% of referrals went to ten domains, while over 20% went to Google.

Source: Semrush, April 7, 2026.

There is an important qualification: the study examined a US clickstream panel, not every ChatGPT user worldwide.

It also found that ChatGPT activated its web-search feature on approximately 34.5% of queries in February 2026, compared with 46% in late 2024.

This suggests that many ChatGPT responses still do not involve a live web search.

The difference between a chatbot answer and a search-generated answer is therefore not merely cosmetic. The system may draw on previously learned information, external web results, or a combination of both.

For publishers, this distinction matters because a live web search creates an opportunity for a website to be cited and visited. A response based primarily on existing model knowledge may not produce the same opportunity.

4. How People Searched Before AI Chatbots

To understand why the current changes matter, it helps to remember how the internet worked before conversational AI became mainstream.

For much of the internet's history, search was built around a relatively simple relationship.

A user entered a query. A search engine retrieved relevant documents. An algorithm ranked those documents. The user selected a result.

Google became particularly successful because it made this process fast and useful across an enormous collection of websites.

Instead of asking a librarian to locate a particular book, people could search a digital library containing billions of pages.

The search engine did not necessarily know the complete answer to every question. Its primary function was to identify relevant information and help users find it.

4.1 The Rise of Keywords and SEO

As search engines became central to online discovery, websites began competing for visibility.

Businesses created pages targeting particular search terms. Newspapers optimised headlines. Bloggers researched keywords. Online stores improved product descriptions.

This created an entire industry around search engine optimisation, commonly known as SEO.

A website might publish a guide titled Best Budget Smartphones in 2026 because people were searching for that exact information.

The system rewarded pages that appeared relevant, useful and authoritative, although search algorithms also became vulnerable to manipulation.

Over time, many users began encountering pages filled with repetitive content, excessive advertisements, affiliate links and articles written primarily to attract search traffic.

This did not mean that Google stopped providing valuable results. It meant that finding genuinely useful information could require more effort.

A user searching for a simple answer sometimes had to navigate several pages before discovering the information they needed.

4.2 The Shift From Keywords to Conversations

AI chatbots introduced a different interaction.

Instead of typing:

Best laptops students 2026

A user could write:

I am a university student with a budget of $700. I need a laptop for programming, online classes and occasional video editing. Which three models should I consider, and what are their disadvantages?

The system can interpret the context, organise the criteria and provide a comparative explanation.

The user can then ask:

Which one has the best battery life?

Or:

Would the cheaper model still be suitable for programming?

This creates a continuous conversation rather than a sequence of disconnected searches.

It resembles the difference between reading a catalogue and speaking to an assistant who understands what you are trying to accomplish.

That is one of the fundamental reasons AI search is attracting attention.

5. Why Are Users Switching to AI Search?

The strongest explanation is not that AI has made traditional search useless.

It is that AI has reduced the effort required to reach an initial understanding of a subject.

A particularly useful study was published by the Reuters Institute for the Study of Journalism on October 29, 2026.

Its Generative AI and News Report 2026 examined how people use AI across nine countries: Argentina, Australia, Denmark, Finland, France, Japan, Norway, the United Kingdom and the United States.

The survey found that 51% of respondents across the six countries with comparable previous-year data used standalone generative AI systems weekly in 2026, up from 34% in 2025.

It also found that ChatGPT remained the leading chatbot, with average weekly use across those six countries increasing from 22% to 34%.

Source: Reuters Institute, October 29, 2026.

The researchers also investigated why people use AI to retrieve information rather than relying on other sources.

Their findings are revealing.

Reason for using AI	Share of information-retrieval users
Faster and easier answers	64%
Natural-language interaction	56%
Bringing together large amounts of information	54%
More reliable information than other sources	21%

These figures are from the Reuters Institute's 2026 report. Respondents could select multiple reasons.

Notice something important: reliability was not the leading motivation.

Most users are not necessarily choosing AI because they believe it is more accurate than every website.

They are choosing it because it is convenient.

5.1 Convenience Changes Behaviour

Imagine trying to understand a complicated financial concept.

The traditional process might involve opening several articles, comparing explanations and looking for examples.

An AI assistant can explain the concept, provide a simple example, answer a follow-up question and adjust its explanation when the user does not understand.

This is especially valuable for people who are learning something unfamiliar.

The same pattern applies to travel planning, technology comparisons, language learning, cooking and everyday problem-solving.

5.2 People Can Ask Questions Naturally

Traditional search encourages people to compress their thoughts into short phrases.

Conversational AI allows them to describe a situation in ordinary language.

This can make research more accessible to people who do not know the correct technical vocabulary.

Someone unfamiliar with computer hardware might ask:

Why does my computer become slow when I open several applications?

They do not need to know whether the problem involves RAM, processor utilisation, storage performance or background processes.

The assistant can introduce those concepts as part of its explanation.

5.3 Research Becomes Interactive

A traditional search result is generally designed to answer the query that was entered.

An AI assistant can continue exploring the subject.

It can explain an unfamiliar term, compare two positions, simplify a technical paragraph or organise a long explanation into steps.

The Reuters Institute's 2026 findings suggest that information retrieval is becoming one of the central reasons people use conversational AI. Across the six countries with comparable data, weekly use of AI for getting information increased from 11% in 2024 to 24% in 2025 and 38% in 2026.

This is evidence of a behavioural change, not proof that every one of those information requests has replaced a Google search.

Nevertheless, it demonstrates that AI is moving beyond experimentation and becoming part of everyday information habits.

6. Google's Response: Turning Search Into an AI Conversation

Google understands that the interface through which people find information is changing.

Its response has not been to abandon the search engine that built its business.

Instead, Google has been incorporating AI into the existing search experience.

Three developments are particularly important: AI Overviews, AI Mode and Gemini.

6.1 AI Overviews: Answers at the Top of Google

AI Overviews provide AI-generated summaries above or alongside traditional search results.

Instead of requiring a user to open several websites, Google can present a combined explanation with links to supporting sources.

Google introduced AI Overviews as a major part of its search transformation and has continued expanding the feature.

In its June 3, 2026 investor presentation, Alphabet reported that AI Overviews had more than 2.5 billion monthly users.

Source: Alphabet investor presentation, June 3, 2026.

That number illustrates an important competitive advantage.

Google does not need to persuade billions of existing users to abandon its search engine and download a separate AI application.

It can introduce conversational and generative features directly into an interface people already use.

6.2 AI Mode: A Conversational Search Experience

AI Mode goes further.

Instead of providing only a summary, it allows users to explore a topic through follow-up questions and increasingly detailed responses.

Google described its AI Mode experience in a May 19, 2026 update.

The company reported that AI Mode had surpassed one billion monthly active users globally and that its query volume had more than doubled every quarter since launch.

Google also said that the average AI Mode search was approximately three times as long as a traditional Search query.

Source: Google, May 19, 2026.

This is significant because it shows that the AI revolution is changing Google's own search behaviour.

People are not simply transferring their old short queries into a new interface. They are asking more detailed questions and exploring subjects through longer conversations.

6.3 Gemini: Google's Separate AI Assistant

Google's Gemini is another major competitor to ChatGPT.

Unlike a traditional search page, Gemini is designed to handle conversations, analyse information, work with images and other media, and assist with practical tasks.

On August 11, 2026, Google announced that the Gemini app had surpassed one billion monthly active users.

Source: Google, August 11, 2026.

These figures must not be added together.

Someone can use Google Search, encounter AI Overviews, open AI Mode and use the Gemini application during the same month.

The products have overlapping audiences, and their monthly active user counts measure different experiences.

Nevertheless, the numbers demonstrate that Google is building a substantial AI ecosystem around its existing search infrastructure.

6.4 Google Is Not Simply Defending Old Search

Google's strategy is also about creating new searches.

For example, a user who once searched for a hotel might now ask AI Mode to compare destinations, travel dates, accommodation, transport and prices.

That is a more complex interaction.

Google has argued that AI features are encouraging users to ask more questions rather than reducing their overall search activity.

Its October 28, 2026 earnings statement, reporting the fastest annual query growth in five years, supports the company's position that AI can expand the search market rather than merely redistribute it.

The unanswered question is how much of that growth will translate into website visits and advertising revenue.

More searches do not automatically mean more clicks.

And more clicks do not automatically mean more value for publishers.

7. The Other Companies Competing for the Future of Search

Google and OpenAI are not the only companies trying to change how people find information.

Several other technology companies are building AI-powered search experiences, each with a different advantage.

7.1 Perplexity: Search Designed Around Answers

Perplexity is one of the clearest examples of a company built specifically around AI search.

Its approach is slightly different from ChatGPT's broader assistant model.

Perplexity emphasises research, live information retrieval and references. A user asks a question, and the platform searches for relevant information before producing a structured response with supporting sources.

Its design attempts to combine the convenience of a chatbot with the transparency of a conventional search engine.

In June 2025, Perplexity CEO Aravind Srinivas disclosed that the platform had processed approximately 780 million queries during May 2025. He also said that query volume was growing by more than 20% month over month at that time. The company had a target of reaching one billion queries per week by the end of 2025.

Source: Search Engine Land, June 6, 2025.

However, that target should not be confused with a confirmed achievement.

As of November 5, 2026, a newer, independently verified companywide query total using the same disclosed methodology has not been established by the sources examined for this article.

The important development is not simply Perplexity's volume. It is the product concept: a search engine that begins with an answer and provides references rather than making the user assemble everything from a list of links.

7.2 Microsoft Copilot and Bing

Microsoft has an interesting position in this competition.

Its Bing search engine has historically been Google's largest conventional competitor in many international markets.

Statcounter reported that Bing held approximately 4.8% of worldwide search-engine referral share in October 2026.

Microsoft has also incorporated its Copilot technology into Bing.

Its current AI search experience includes two important formats:

• AI Overview: A generated summary at the top of search results.
• AI Mode: A conversational experience that supports follow-up questions and deeper exploration.

Microsoft's documentation explains that these experiences can display references and links, allowing users to investigate the material behind an answer.

Source: Microsoft Bing AI.

Copilot also benefits from Microsoft's existing distribution through Windows, Edge, Microsoft 365 and enterprise software.

That is a different competitive advantage from OpenAI's rapidly growing standalone chatbot audience.

A person might begin researching something through Bing, continue the discussion in Copilot and then use an AI feature inside Word or Excel.

The distinction between searching and completing a task is becoming less obvious.

7.3 Apple: Bringing AI Search Into the Operating System

Apple has a particularly interesting opportunity.

Its advantage is not necessarily building the most popular independent search website.

It is controlling the operating systems and devices through which millions of people interact with information.

On September 14, 2026, Apple announced the next generation of Apple Intelligence, introducing a substantially redesigned conversational assistant called Siri AI.

The new experience is designed to understand personal context, answer questions across a wide range of subjects and interact with information in a user's messages, emails, photographs and applications.

Apple said Siri AI would begin rolling out in beta with its 2027 software releases, initially in English, with additional languages planned.

Source: Apple, September 14, 2026.

This is important because search may increasingly happen without someone deliberately opening a search engine.

A user might simply ask a phone:

Find the email containing my hotel reservation and tell me how far the hotel is from the airport.

That is information retrieval combined with an action.

Apple's approach illustrates how AI assistants could become a layer between users and the applications they already depend on.

7.4 Meta AI: Search Inside Social Platforms

Meta has another advantage: distribution.

Its AI assistant is available through Meta AI, WhatsApp, Instagram, Messenger and Facebook.

Rather than requiring users to adopt an entirely new information platform, Meta can introduce AI assistance inside applications that people already use.

Meta's current product information describes an assistant capable of answering questions, helping with research, generating images and using real-time web information.

Source: Meta AI official product information.

Pew Research Center's February 17–23, 2026 survey found that 14% of US adults reported having used Meta AI. The corresponding figures were 44% for ChatGPT, 24% for Gemini and 17% for Copilot. These were measures of reported usage, not weekly activity or search share.

Source: Pew Research Center, June 17, 2026.

Meta's wider family of applications also gives it an enormous potential distribution network, although the total number of people using Facebook, Instagram or WhatsApp must not be confused with the number using Meta AI.

7.5 A Competition With Different Strengths

The competitive picture is becoming more complicated than a simple Google-versus-ChatGPT comparison.

Company	Main advantage	Search strategy
OpenAI	Large conversational AI audience	ChatGPT Search and interactive answers
Google	Search infrastructure and existing distribution	AI Overviews, AI Mode and Gemini
Microsoft	Windows, Bing and enterprise software	Copilot-powered Bing experiences
Perplexity	Research-oriented search experience	Answers with references
Apple	Hardware and operating-system integration	Siri AI and Apple Intelligence
Meta	Social and messaging distribution	Meta AI inside existing applications

There is no single universally accepted measurement that compares the search activity of all these products.

Some publish monthly active users. Others disclose selected query milestones. Some are measured through third-party traffic estimates.

The figures should remain separate until comparable data becomes available.

8. What Happens to Publishers and Websites When AI Answers Replace Clicks?

This is arguably the most important part of the entire search revolution.

Search engines did not simply help people find information. They helped create the economic system that supported a large part of the modern internet.

A publisher could write an article, rank for a search query, attract visitors and earn money through advertising, subscriptions, affiliate links or product sales.

AI-generated answers introduce a different relationship.

A chatbot can use information from several websites to produce one response. The reader receives the explanation without necessarily visiting the websites that contributed to it.

That creates a difficult question:

If AI systems increasingly answer questions using information published by websites, how will those websites continue earning enough to produce that information?

8.1 The Rise of Zero-Click Searches

A zero-click search occurs when a person receives enough information from a search-results page without opening an external website.

This behaviour existed long before generative AI.

A person searching for the weather, a mathematical calculation or the opening time of a shop might receive the answer directly from Google.

AI summaries make this behaviour possible for more complicated questions.

Pew Research Center published an important study on July 22, 2025, examining how people interacted with Google's AI-generated summaries.

The research analysed 68,879 unique Google searches performed during March 2025 by 900 US adults who shared their browsing activity.

Approximately 18% of the searches in the dataset produced an AI summary.

Source: Pew Research Center, July 22, 2025.

The researchers found:

User behaviour	AI summary present	No AI summary
Clicked a traditional search result	8%	15%
Ended browsing after search	26%	16%

Only approximately 1% of visits to pages containing an AI summary resulted in a click on a source link within that summary.

This suggests that AI summaries can satisfy a significant proportion of information needs without sending visitors to the websites that supplied the underlying material.

However, the study has limitations.

It examined a particular group of US users, a single month of browsing activity and the search results collected during April 2025. It does not prove that every AI search feature will produce the same behaviour.

Nevertheless, it provides measurable evidence that the appearance of an AI answer can change what users do next.

8.2 Publishers Are Reporting Declining Search Referrals

The concerns are not theoretical.

The Reuters Institute's 2026 Digital News Report discussed Chartbeat data covering more than 2,500 websites.

According to that analysis, Google organic search traffic to those sites declined by approximately 33% worldwide between November 2024 and November 2025. The decline in the United States was approximately 38%.

Source: Reuters Institute publisher traffic analysis.

A separate Reuters Institute article published on October 13, 2026, reported that Google Search referrals to hundreds of publishers had declined by approximately 40% between July 2025 and July 2026, according to Chartbeat's annual publisher analysis.

These are different datasets and time windows, so their percentages should not be combined.

It is also important not to attribute every decline directly to AI.

Other factors include:

• Changes in Google's ranking systems.
• Changes in user behaviour.
• Competition from social media and video platforms.
• Changes in news consumption.
• Seasonal differences and industry-specific trends.
• Changes in the types of content people search for.

Nevertheless, AI-powered answers are an increasingly important part of the explanation publishers must investigate.

8.3 The Difference Between Being Cited and Receiving a Visitor

Imagine THE JORNAL publishes a detailed article explaining a new scientific discovery.

A reader asks ChatGPT about the discovery.

The assistant finds the article, extracts a useful explanation and cites THE JORNAL as one of its references.

That is valuable visibility.

But the reader might never visit the original article.

The publisher receives recognition, but not necessarily a website session.

This creates two different measures of success:

1. AI visibility: How frequently a publication appears in AI-generated responses and recommendations.
2. Referral traffic: How many people actually visit the publication after encountering an AI response.

A publication could perform well in the first measurement and poorly in the second.

This is why simply counting AI citations is not enough to understand whether AI search is commercially beneficial to publishers.

8.4 What About Independent Blogs?

Smaller websites may face particularly difficult choices.

An independent blog can spend hours researching and writing a detailed article, only to discover that an AI assistant can summarise the essential information in a few paragraphs.

The challenge is not limited to large newspapers.

It affects:

• Educational websites.
• Technology blogs.
• Science publications.
• Travel guides.
• Financial advice websites.
• Product comparison websites.
• Independent research publishers.

The more easily an article can be reduced to a simple factual answer, the more vulnerable its traffic may become to answer engines.

However, this does not mean that all written content will lose value.

Original investigations, expert analysis, exclusive interviews, detailed experiments, photographs, first-hand experience and specialised reporting provide value that cannot always be replaced by a short summary.

The strongest publishers will need to offer reasons for readers to visit the original source rather than simply receive its conclusions indirectly.

9. How AI Search Is Changing Advertising and Online Business

Search advertising has historically been one of the most valuable digital advertising models.

The reason is straightforward.

A person searching for a particular product often demonstrates clear purchasing intent.

Someone searching for best washing machine under $500 may be much closer to buying than someone casually scrolling through social media.

Google built a highly successful advertising system around this behaviour.

AI search introduces new possibilities.

9.1 Advertising Inside Conversational Answers

A conversational assistant can understand more about what a user wants.

For example:

I need a comfortable office chair for eight hours of work every day. My budget is $250, and I have limited space.

This is more detailed than a short keyword query.

It provides information about the user's budget, priorities and intended use.

Advertising systems may use this context to present relevant products and services.

Google has already been expanding advertising in AI Overviews and testing advertisements in AI Mode. Its May 21, 2025 announcement described how advertisements could appear within or alongside AI-generated search experiences.

Source: Google advertising and AI search information.

OpenAI is also developing advertising as a commercial part of ChatGPT.

On August 31, 2026, OpenAI announced that ChatGPT advertising had reached a $1 billion annualised revenue run rate in less than 200 days after launch, with tens of thousands of advertisers using the platform.

Source: OpenAI, August 31, 2026.

An annualised revenue run rate is not the same as $1 billion already collected over a completed year. It is a measure of the pace at which the business was operating at that time.

Still, the announcement demonstrates that AI assistants are becoming advertising destinations in their own right.

9.2 Advertisers Will Need to Understand Intent

Traditional search advertising often targets keywords.

AI advertising may increasingly target the underlying task.

For example, a user asking for a family holiday recommendation might receive information about flights, hotels, travel insurance and local activities within one conversation.

That creates opportunities for multiple businesses.

However, it also introduces a challenge: advertisements must be clearly distinguished from independent recommendations.

If a user believes that an AI assistant is recommending the best product objectively, but the recommendation is influenced by commercial placement, transparency becomes essential.

The future success of AI advertising will depend partly on whether platforms can preserve user trust while generating revenue.

9.3 Businesses Will Need More Than Search Rankings

A business might rank highly on Google but rarely appear in AI-generated recommendations.

Another business might receive relatively little traditional search traffic but be repeatedly mentioned in relevant AI answers.

This means companies will increasingly need to understand how their products, services and reputation appear across different discovery platforms.

The objective should not simply be to appear everywhere.

It should be to appear accurately when a customer is looking for something the business genuinely offers.

10. Is Traditional SEO Dead?

No.

But SEO is changing.

For years, the central objective was to make a website visible in traditional search results.

Now publishers must consider two related questions:

• Can search engines discover and rank the page?
• Can AI systems identify, understand and accurately represent the information?

The second question has created growing interest in terms such as Answer Engine Optimisation (AEO) and Generative Engine Optimisation (GEO).

These terms describe efforts to improve the visibility of websites and brands in AI-generated responses.

However, they should not be treated as magic techniques.

There is no guaranteed method for forcing an AI assistant to recommend a particular website.

10.1 What Makes Content Useful to AI Systems?

The most durable approach is to publish genuinely valuable information.

For a website, that means:

• Original research. Publish information that readers cannot easily find repeated elsewhere.
• Clear explanations. Explain technical subjects in language that people can understand.
• Reliable sources. Support important claims with primary research, official documents and credible publications.
• Transparent authorship. Make it clear who created the content and what expertise or experience supports it.
• Accurate dates. Identify when information was published and updated.
• Structured presentation. Use meaningful headings, descriptive paragraphs, tables and properly organised information.
• Technical accessibility. Ensure that important content is crawlable, indexable and accessible without unnecessary technical barriers.

Google's Search Central guidance has also evolved. In May 2026, Google published guidance addressing its generative AI search features, emphasising useful, distinctive content and explaining how established SEO practices continue to matter.

Source: Google Search Central documentation and updates.

The lesson is that good SEO remains important, but ranking alone is no longer the complete picture.

10.2 Why Original Reporting Matters More

Consider two articles about a new scientific development.

The first simply repeats a press release.

The second examines the original research, explains the experimental method, discusses the limitations, interviews researchers and provides independent context.

Both may contain similar basic facts.

But the second provides considerably more value.

In an environment where AI can summarise existing information, originality becomes an important competitive advantage.

A publication that consistently produces trustworthy original reporting gives readers a reason to visit directly.

It also gives other information systems more distinctive material to discover and cite.

10.3 What Publishers Should Measure

Website owners should not abandon their existing analytics.

Instead, they should expand what they monitor.

Important measurements include:

• Google organic impressions and clicks.
• Search referral traffic.
• AI-generated referral traffic.
• Direct visits.
• Returning readers.
• Newsletter subscriptions.
• Conversion rates.
• Pages and brands mentioned by AI assistants.
• The accuracy of AI-generated descriptions of their content.

A fall in Google clicks does not automatically mean that the audience has disappeared.

Some readers may be encountering a publication through AI answers, newsletters, social platforms or direct visits.

The challenge is to understand where the audience is going and whether those discovery channels produce sustainable value.

11. What This Means for Everyday Internet Users

For ordinary readers, the change is both exciting and complicated.

AI search can make information easier to understand, especially when someone is exploring an unfamiliar subject.

A student can request an explanation of a difficult scientific concept. A parent can compare educational options. A traveller can plan a trip. A small-business owner can investigate competitors and organise a marketing strategy.

The ability to ask follow-up questions is particularly useful.

Instead of opening a new search page whenever something is unclear, the user can continue the discussion.

However, the convenience of AI answers creates a risk: people may become less likely to examine the original evidence.

11.1 The Danger of a Convincing but Incorrect Answer

An AI assistant can produce an explanation that looks professional, sounds confident and contains detailed examples.

That does not guarantee that the information is correct.

AI systems can:

• Misinterpret scientific studies.
• Combine information from unrelated sources.
• Present outdated figures as current.
• Invent references or misrepresent what a source says.
• Miss important disagreements between experts.
• Produce plausible but incorrect explanations.

The problem is particularly serious when a question concerns medicine, financial decisions, legal obligations or public safety.

A traditional search engine gives users a list of sources to examine. AI can make the process easier by synthesising information, but it also places more responsibility on the reader to verify the synthesis.

11.2 How to Check Information From ChatGPT and Other AI Search Tools

Readers do not need to reject AI to use it responsibly.

A few practical habits can make a substantial difference.

1. Ask for the original sources. When an AI assistant gives an important factual answer, ask it to identify the original research, official announcement or document supporting the claim.
2. Open the references. Do not assume that a citation proves the statement is correct. Read the relevant passage yourself.
3. Check publication dates. A report from several years ago may no longer describe the current situation.
4. Compare independent sources. For major claims, compare at least two credible sources that do not simply repeat the same original announcement.
5. Distinguish facts from estimates. If an AI system gives a percentage, ask whether it comes from an official measurement, an independent study or a mathematical projection.
6. Look for disagreement. If credible sources report different figures, investigate their definitions and methodologies rather than automatically accepting the largest number.
7. Visit specialist websites when necessary. For scientific research, consult original papers. For financial information, consult official financial documents. For government decisions, check the relevant government publication.

The best approach is to use AI as an assistant for understanding information, not as a replacement for independent judgement.

12. What Happens Next? Three Possible Directions for Search

Nobody can say exactly how the search industry will look five years from now.

However, current developments point towards several important possibilities.

12.1 Search Engines and AI Assistants Will Become More Similar

The distinction between a search engine and a chatbot is already becoming less obvious.

Google offers conversational AI.

ChatGPT offers web search.

Bing combines traditional results with Copilot-powered answers.

Perplexity provides search results in a conversational format.

Apple is integrating AI into its operating system.

These products are gradually adopting similar capabilities, even though their business models and technical approaches remain different.

The likely result is not necessarily the disappearance of search engines.

It may be the transformation of search into a more conversational experience across several competing platforms.

12.2 Search May Become Task-Oriented

Today's search often begins with a question.

Tomorrow's search may begin with a goal.

Imagine someone planning to purchase a new computer.

Instead of asking several disconnected questions, the user could tell an AI assistant:

Find three laptops suitable for programming, compare their prices, check independent reviews, and identify the best option within my budget.

The assistant could then retrieve information, compare products, identify weaknesses and help the user make a decision.

This is a shift from finding documents to completing research tasks.

The same idea could apply to travel planning, education, professional research and business operations.

It also explains why companies are investing in AI agents that can interact with applications and perform multi-step activities.

12.3 The Internet May Become Less Dependent on a Single Gateway

For many years, Google occupied a particularly powerful position between users and websites.

Businesses built strategies around its search results.

Publishers depended on its referrals.

Advertisers invested heavily in its advertising ecosystem.

AI introduces additional gateways.

A person might discover a website through ChatGPT, Gemini, Perplexity, Copilot, Meta AI or a recommendation from another assistant.

This could create a more distributed discovery environment.

However, it could also create new forms of concentration.

If a handful of AI platforms become responsible for selecting which websites are mentioned in answers, those platforms could gain considerable influence over what information people encounter.

The internet might have more discovery channels while becoming more dependent on a smaller number of AI intermediaries.

That is one of the central issues that publishers, regulators and technology companies will need to address.

13. What We Know, and What We Still Do Not Know

The current evidence supports a major change in information discovery, but it does not justify declaring that Google has been replaced.

It is useful to separate the strongest findings from the questions that remain unresolved.

What is established	What remains uncertain
OpenAI reported 1.2 billion weekly ChatGPT users on November 3, 2026.	The precise proportion of those users who use ChatGPT primarily for search.
OpenAI's consumer product lead reported search-like query volume at approximately 20% of Google's.	Whether both companies use identical counting methods.
Google disclosed more than five trillion annual searches in May 2025.	Google's exact current worldwide annual query total.
Google reported its fastest year-over-year search-query growth in five years on October 28, 2026.	How much of that growth comes specifically from AI features.
Statcounter recorded approximately 91.05% Google search-engine referral share in October 2026.	The complete market share of all conversational AI search activity.
Independent studies show that AI-generated summaries can reduce traditional search-result clicks.	How much of publishers' traffic decline is directly caused by AI.
AI platforms are increasingly sending visitors to external websites.	Whether those referrals can replace traditional search traffic and revenue.

There are several reasons why the distinction matters.

First, ChatGPT's 20% figure is a comparison of search-like query volume, not a direct market-share measurement.

Second, Google is continuing to grow its search activity, according to its latest October 2026 earnings announcement.

Third, website traffic is a different measurement from searches and queries.

Fourth, the future economics of AI search remain unsettled. The companies must demonstrate that their advertising, subscription and other business models can support the enormous infrastructure costs involved.

Finally, the impact on publishers will vary considerably by industry. A news organisation, a scientific journal, a product comparison website and a small travel blog will not necessarily experience the same changes.

The next few years will provide better evidence.

Conclusion: The Internet Is Moving From Finding Answers to Understanding Them

The rise of ChatGPT represents more than the arrival of another search competitor.

It challenges an assumption that has shaped the internet for decades: that finding information should begin with a list of documents.

For years, Google helped users navigate the enormous collection of information published online.

Now, AI assistants are attempting to perform part of that navigation themselves.

They can interpret questions, combine information, explain unfamiliar concepts and maintain a conversation while users explore a subject.

The reported 20% comparison between ChatGPT's search-like query volume and Google's is therefore an important milestone. The reported 1.2 billion weekly users demonstrate how widely the platform has spread.

But Google remains enormously powerful.

Its approximately 91% October 2026 search-engine referral share, its five-trillion annual search benchmark and its continued query growth demonstrate that the traditional search ecosystem has not disappeared.

Indeed, Google itself is becoming an AI search platform.

The more immediate transformation may be happening in the relationship between readers and publishers.

When users receive complete answers without visiting original websites, the economics of online publishing become more complicated. Visibility, citations, audience relationships and original reporting become increasingly important.

For readers, the opportunity is greater access to explanations and more natural ways of exploring information.

For publishers, the challenge is to remain valuable when answers can be delivered without a visit.

And for technology companies, the competition is no longer simply about which platform can display the most relevant links.

It is about which platform people trust to help them understand the world.

Google built much of the modern internet around finding information. AI is now attempting to build the next version around understanding it.

The transition is already underway, but its final destination remains uncertain.

Frequently Asked Questions (FAQ)

1. Is ChatGPT Replacing Google?

No. ChatGPT is becoming a significant alternative for information discovery, research and conversational questions, but Google remains the dominant conventional search engine. Google's approximately 91.05% worldwide search-engine referral share in October 2026 demonstrates that traditional search remains extremely powerful. The more likely development is that people will use both services for different purposes rather than completely abandoning one for the other.

2. How Big Is ChatGPT's Search Share?

OpenAI's consumer product leadership reported on November 3, 2026, that ChatGPT handles approximately 20% of Google's search-like query volume. However, an independent seoClarity model published on October 28, 2026, estimated a lower figure of approximately 7.5%. The difference comes from different definitions and counting methods. Neither figure should be confused with conventional search-engine market share or website referral traffic.

3. Does ChatGPT Really Have 1.2 Billion Weekly Users?

OpenAI's consumer product leadership reported approximately 1.2 billion weekly active ChatGPT users in November 2026. This measures people actively using ChatGPT during a week, not people who necessarily use it as their search engine. Users may also use the platform for writing, programming, translation, education, research, image generation and many other activities.

4. Is Google Losing Users to ChatGPT?

Some people are choosing ChatGPT and other AI assistants for questions they previously might have searched on Google. However, there is no conclusive evidence that Google is experiencing an overall collapse in search usage. On October 28, 2026, Google reported that its year-over-year search-query growth rate was the highest in five years. The competition may be expanding the total number of information searches while changing where some of those searches take place.

5. Are AI Search Answers Accurate?

AI search answers can be useful and accurate, particularly when they are supported by reliable, current sources. However, they can also contain outdated figures, incorrect explanations, misleading summaries or fabricated references. Accuracy depends on the subject, the information available and the quality of the system's sources. Important claims should always be checked against original documents or independent authoritative sources.

6. Should I Trust ChatGPT for News, Health or Financial Information?

ChatGPT can help explain news reports, medical terminology and financial concepts, but it should not be treated as the final authority. For breaking news, check original reporting and official statements. For health questions, consult qualified healthcare professionals and established medical sources. For financial decisions, examine official documents and seek appropriate professional advice when necessary. AI should support informed decisions rather than replace expert judgement.

7. What Is Zero-Click Search?

A zero-click search occurs when someone receives the information they need directly from a search-results page or AI-generated answer without visiting an external website. For example, Google might display an AI summary explaining a scientific concept, allowing the user to understand the basic answer without opening the original article. This can reduce referral traffic for publishers, even when their information contributes to the answer.

8. What Does AI Search Mean for Bloggers and Publishers?

Publishers may receive fewer visitors from traditional search results when AI systems provide answers directly. However, they can adapt by publishing original research, expert analysis, exclusive reporting, useful photographs, detailed explanations and trustworthy information. They should also monitor AI referrals, search visibility, direct traffic, returning visitors and subscriptions. The objective is to build an audience that values the original publication, not simply to appear in search results.

9. What Should Businesses Do to Adapt to AI Search?

Businesses should maintain strong traditional SEO while improving their visibility across AI-powered discovery platforms. This includes maintaining accurate business information, publishing helpful product and service descriptions, collecting genuine customer reviews, providing clear answers to common questions and developing a trustworthy online reputation. Companies should also measure whether AI referrals produce meaningful enquiries, sales or conversions rather than focusing only on mentions.

10. Will Google Disappear Because of AI?

There is currently no strong evidence that Google will disappear. Its search infrastructure, global distribution, advertising business, Android ecosystem, browser integration and enormous existing audience provide significant competitive advantages. Google is also incorporating AI into its own products through AI Overviews, AI Mode and Gemini. The more likely possibility is that Google will continue changing what a search engine looks like.

11. How Can I Check Whether an AI Answer Is Correct?

Ask the AI assistant to provide its original sources, open those references and read the relevant sections. Check publication dates, compare independent sources and distinguish confirmed facts from estimates or predictions. For important scientific claims, look for original research. For public announcements, consult official websites. If sources disagree, investigate their methodologies instead of accepting a single answer without examination.

12. Is Traditional SEO Still Worth Doing in 2026?

Yes. Traditional SEO remains important because search engines continue to send substantial traffic to websites. However, publishers should expand their strategy to include AI visibility, original reporting, structured information, reliable citations, author transparency and strong audience relationships. There is no guaranteed formula for appearing in AI answers, but useful and trustworthy content remains a sensible long-term investment.

13. What Is the Difference Between Search Share and Website Traffic Share?

Search share describes a platform's portion of measured search activity, depending on the methodology being used. Website traffic share describes the visits that a platform sends to external websites. A person can ask ChatGPT ten research questions and receive answers without opening any websites. Those questions may contribute to AI search activity, but they do not necessarily create ten external website visits. This is why the two measurements must never be treated as interchangeable.

14. Which AI Search Engine Is Best?

There is no universally best platform for every task. ChatGPT is useful for conversational research and multi-step explanations. Google remains powerful for discovering websites, local information and conventional search results. Perplexity emphasises referenced research. Microsoft Copilot connects AI assistance with Microsoft's software ecosystem, while Gemini integrates with Google's services. The best choice depends on the question, the need for current information and the quality of the sources provided.

Sources and Further Reading

The following sources support the principal statistics and developments discussed in this article. Dates are included to distinguish current measurements from historical benchmarks.

OpenAI and ChatGPT

1. Big Technology – OpenAI: ChatGPT Search Is About 20% of Google's — November 3, 2026. Adam Fry's interview and the 1.2 billion weekly active users disclosure.
2. The Verge – OpenAI says ChatGPT is catching up to Google in search — November 3, 2026. Explanation of OpenAI's search-like query comparison.
3. OpenAI – A milestone in expanding access to AI — August 31, 2026. More than one billion weekly active users and ChatGPT advertising developments.

Google and Conventional Search

4. Google – More opportunities for your business on Google Search — May 21, 2025. Google's disclosure of more than five trillion annual searches.
5. Alphabet – Q3 2026 earnings call: Remarks from our CEO — October 28, 2026. Google's latest statement about search-query growth.
6. Statcounter – Worldwide Search Engine Market Share — October 2026 data. Google, Bing and other conventional search-engine referral shares.
7. Statcounter – Methodology and Frequently Asked Questions — Explanation of how search referrals and market-share statistics are measured.
8. Google – How AI Mode is changing and expanding the way people search — May 19, 2026. AI Mode adoption and search behaviour.
9. Alphabet – Investor Presentation, June 2026 — June 3, 2026. AI Overviews' reported monthly audience.
10. Google – More than 1 billion people are using the Gemini app every month — August 11, 2026. Gemini's monthly audience.

Independent Research and Industry Measurements

11. seoClarity – ChatGPT vs. Google in 2026: Updated Search Volume Estimate — October 28, 2026. The alternative 7.5% model and its assumptions.
12. Semrush – ChatGPT Traffic Analysis: Insights from 17 Months of Clickstream Data — April 7, 2026. ChatGPT referral traffic, user journeys and live search estimates.
13. Search Engine Land – ChatGPT Sends Less Than 1% of Google's Traffic — October 21, 2026. The 100-website referral traffic study.
14. Pew Research Center – Google Users Are Less Likely to Click on Links When an AI Summary Appears — July 22, 2025. Browsing behaviour and AI summary click-through data.
15. Pew Research Center – Americans and AI 2026 — June 17, 2026. AI chatbot adoption and usage patterns in the United States.
16. Reuters Institute – Generative AI and News Report 2026 — October 29, 2026. International research on AI use, information discovery and public behaviour.
17. Reuters Institute – Life After SEO: How Newsrooms Are Adapting to the AI Search Shift — October 13, 2026. Publisher traffic and changing search visibility.

Other AI Search Competitors

18. Search Engine Land – Perplexity Grows to 780 Million Monthly Queries — June 6, 2025. Perplexity's CEO-reported query milestone.
19. Microsoft – Search with AI — Bing AI Overview and AI Mode product information.
20. Apple – Siri AI: A More Capable and Personal Assistant — September 14, 2026. Apple's next-generation Siri announcement.
21. Meta AI – Official Product Information — Meta AI features and availability across Meta applications.
22. Google Search Central – Latest Documentation Updates — Search documentation and guidance for website owners, including generative AI features.

Editorial note: Statistics in this article refer to different reporting periods and measurement methodologies. Company-reported figures, third-party estimates and observed referral traffic are identified separately wherever possible. They should not be interpreted as directly interchangeable measures of search-market share.

THE JORNAL — New ideas. Wider horizons.