When a report about AI chatbots and the news crossed my desk, the easy thing was to read it as a media story and scroll on. It isn't only a media story. Buried in the Reuters Institute's Digital News Report 2026 is a picture of how people now ask a question, get an answer, and stop. They don't click through to where the answer came from. That same behavior is starting to decide whether a small business gets found at all.
For a busy owner, the lede is this: more and more, people get the answer they wanted inside the chatbot and never visit the source. If your customers are doing that with the news, they are starting to do it when they look for a service nearby, like someone to fix a furnace or a bookkeeper for a small agency. The page that used to earn the click is being read aloud by the chatbot instead, and the business that isn't part of that answer never knows it was passed over.
What the report found
The Reuters Institute for the Study of Journalism at the University of Oxford published its Digital News Report 2026 on Tuesday, June 16, 2026. The chapter on AI chatbots was written by Dr Amy Ross Arguedas. The full report draws on almost 100,000 online news consumers across 48 markets, fielded by YouGov, so this isn't a small sample or a single-country quirk.
A few numbers tell the story:
- Weekly use of an AI chatbot for news rose to 10% on average across the markets surveyed, up from 7% the year before.
- Only 1% of people name AI as their main source of news. Chatbots are still a supplement, not a replacement, for most people.
- The audience skews young and engaged. People aged 18 to 24 are three times more likely than those over 55 to use chatbots for news. In the youngest group, 17% use chatbots for news, against 5% of the oldest. The 25-to-34 group grew the most, up 4 percentage points from 2025.
So adoption is climbing, it's led by younger and more curious people, and almost nobody has gone all-in. That's the early stage of a habit forming, which is exactly when a business has time to get ready.
People ask the chatbot, then stay put
The most useful finding for a business owner has nothing to do with journalism. It's about clicking.
Across all respondents, only 4% say they "always or often" click through to the original source from an AI answer. Compare that to 19% from search and 17% from social media. Even among people who actively use chatbots, 51% say they click through wanting more detail, against 59% for search users and 60% for social users.
of people "always or often" click through to the original source from an AI chatbot answer, against 19% from search and 17% from social media.
Source: Reuters Institute Digital News Report, 2026Read that again with your own customers in mind. When someone uses Google or scrolls social, they're far more likely to tap through to a website. When they ask a chatbot, the answer usually settles the question right there. As Arguedas put it, "AI is not simply another route to headlines." It's a different behavior, and it ends inside the chat window.
AI is not simply another route to headlines.
For a business, the consequence is quiet and easy to miss. If a buyer asks ChatGPT or Google Gemini for a recommendation and the chatbot answers without sending them anywhere, the business that wasn't named in that answer was invisible, and there's no click in the analytics to show it happened. You can't measure a visit that never occurred.
Users are doing genuine evaluation
It would be a mistake to picture chatbot users as headline-grabbers. The report shows they use these tools in layered ways. Among the uses people report:
- Asking follow-up questions: 42%
- Getting the latest news: 35%
- Summarizing: 34%
- Assessing how reliable a source is: 33%
- Making something easier to understand: 30%
A third of users are asking the chatbot to judge whether a source can be trusted. They're doing genuine evaluation, not lazy skimming. Arguedas describes the draw plainly: "Part of the appeal of chatbots lies in their ability to provide highly personalised, low-effort responses at scale."
That's the behavior to picture for your own buyers. They'll ask the chatbot who does the work, then ask a follow-up about reviews, price, or whether you're legit, all without leaving the chat. The question for a small business becomes simpler than it sounds: when a chatbot assembles that answer, does it have a clear, accurate, trustworthy picture of you to draw from?
Why this concerns a small business now, not later
There's a longer-term pattern the report names directly. Arguedas writes that "growing use of AI chatbots for news will increasingly eat into referral traffic as users get more detailed, personalised answers to their questions within the chatbot environment."
Swap "news" for "local services" or "products" and you have the trend worth watching. The click-through web that businesses have optimized for over the last decade still works, but a slice of those questions is being answered inside chatbots that send no one anywhere. The 1% who treat AI as their main source today is small. The 10% weekly users, the 3x lead among the young, the year-over-year growth: those are the early lines of a curve, and the smart time to learn the new behavior is while it's still forming.
Getting ready costs attention, not budget. The work is understanding how answers get built inside a chatbot, not only how pages rank on Google.
Three things you can do this week
Small, concrete, no IT department required.
1. Ask the chatbots about yourself
Open ChatGPT and Google Gemini and ask the kind of question a real customer would: who's a good version of your service in your town, or what does your business do? See what comes back. If you're missing, wrong, or thin, that's your first repair list. This takes ten minutes and tells you exactly where you stand.
2. Make your basic facts clear and consistent everywhere
Chatbots assemble answers from what they can find about you. The cleaner and more consistent your name, location, services, hours, and a plain description of who you help, the easier it is for an AI answer to get you right. Check that your website, your Google Business Profile, and your main listings all say the same true things. Consistency is the cheap, high-value work here.
3. Write one plain-language page that answers your top customer questions
The same things people use chatbots for, follow-up questions and reliability checks, are the things a good page can supply: what you do, who you serve, what it costs, why you can be trusted. Write it in clear, direct language a chatbot can quote without guessing. That page does double duty: it helps a human who does click, and it gives the chatbot accurate material to pull from for the human who doesn't.
Where this fits in the bigger picture
Getting found inside a chatbot answer is one piece of a larger climb. In The 7 Levels of AI Proficiency, the early levels are about using the tools yourself with confidence; the middle levels are about understanding how the tools work, which is what lets you ask how a chatbot decides what to say about your business instead of guessing. You don't have to be at the top to act on this report. Asking the chatbots about yourself this week is a concrete step, and it's available to anyone.
The behavior in this study is still early. That's the good part. The owners who learn how answers get assembled now will be the ones a chatbot can describe accurately when a customer asks.
Related reading: Level 4: The Commander (Context Engineer).
Sources
- The emerging uses of AI chatbots for news, and what it means for journalism (Digital News Report 2026)
- Reuters Institute Digital News Report 2026
Frequently Asked Questions
Does this mean my website doesn't count anymore?
No. Search still drives far more click-through than AI does (19% versus 4% who always or often click through), and most people still don't use AI as their main source. A clear, accurate website is exactly what helps a chatbot describe you correctly. The point is to make your site work for both the human who clicks and the chatbot that reads it.
Which chatbot should I focus on?
The Reuters report names ChatGPT and Google Gemini as examples but does not rank which one is used most for news, so there's no single winner to chase. Checking both with a customer's question is a sensible start.
Is this a worldwide trend or just a few countries?
The report covers almost 100,000 people across 48 markets. Adoption varies by place. As one example, in South Korea 8% say they always or often click through to the original source of a chatbot's news answer, higher than the 4% overall. The broad direction, more weekly use and low click-through, shows up across the markets surveyed.
Find your AI Proficiency level
The free 7 Levels assessment places you across seven stages of AI capability. Under ten minutes. Research-backed scoring.