Half now use AI every week, but the UK lags, Oxford survey finds

What did the study find?

The study found that AI assistants have moved from novelty to habit in a single year: half of the public in six countries now uses one every week, and finding information has become the main reason. The Generative AI and News Report 2026, by Felix Simon, Rasmus Kleis Nielsen and Richard Fletcher of the Reuters Institute for the Study of Journalism at the University of Oxford, was published on 29 October 2026.

Five findings matter most for companies that depend on being found by buyers.

• Use has jumped. Weekly use rose from 34% to 51% in a year across the six countries tracked since 2024. Daily use went from 7% in 2024 to 16% in 2025 and 28% in 2026.
• Professionals lead. 64% of people in professional occupations use an AI assistant weekly, well above the average. These are the people who research and recommend suppliers.
• Looking things up is now the main use. 38% used AI to get information in the past week, up from 24% in 2025 and 11% in 2024. Researching a topic, asking for advice and asking for recommendations were among the most common tasks.
• The market is no longer one platform. ChatGPT is still the most used, but Gemini more than doubled, and Copilot and Claude grew from small bases.
• AI answers reach people who never open an assistant. Across all nine countries, 57% said they had seen or engaged with AI-generated answers in search engines in the past week.

Trust is high among those who use these tools: three-quarters of people who used AI to answer factual questions said they trust the answers. Yet only around one in five users said they use AI because the information is more reliable. Convenience is the main draw.

How was the study done?

YouGov surveyed about 2,000 people online in each of nine countries between 14 August and 28 September 2026, and the Reuters Institute compared six of them with the same survey in 2024 and 2025.

The nine countries were Argentina, Australia, Denmark, Finland, France, Japan, Norway, the UK and the US; the six trend countries are Argentina, Denmark, France, Japan, the UK and the US. Samples were set by quotas for age, gender, region, education and political leaning, and weighted.

The limits are stated plainly by the authors. It is an online, non-probability sample, so it under-represents people who are rarely online and has no conventional margin of error. Answers rely on what people recall. The survey covers the general public, not business buyers specifically, and it does not ask about product research or supplier selection. Gemini was asked about as a standalone tool, separate from Google Search. Ireland was not included.

What does it mean for B2B shortlists?

It means the buyer researching your category is increasingly likely to ask an AI assistant before, or instead of, a search engine. Among people who use AI assistants, more said they now use search engines less than said they use them more, by a margin of 13 points.

The study does not measure business buying directly, so this is an inference, but a reasonable one: the professionals who draw up and approve supplier lists are the heaviest users, and their most common tasks are researching, seeking advice and asking for recommendations. Those are the moments a shortlist forms.

The spread of platforms matters as much as the growth. A year ago, measuring ChatGPT alone covered most of the picture. With Gemini more than doubling and Copilot and Claude growing, a company can be recommended on one assistant and missing on another, and the buyer will only see the one they use. Add the AI answers that now sit inside search results, and the first impression of a supplier is often written by AI, not by the supplier.

The trust finding cuts both ways. Users largely trust what they are told, so an accurate, favourable answer carries weight. An out-of-date or wrong answer carries the same weight, and few users go looking for a second opinion.

What does it mean for companies in Ireland and the UK?

For UK companies it means the shift is real but earlier than elsewhere, which is an opening rather than a reason to wait. The UK had the lowest weekly use of the nine countries, at 42%, and only 9% of UK respondents had used an AI agent for a real task in the past week, against 14% across all nine.

A lag in the general public does not mean a lag among buyers. UK companies selling abroad, into the US, the Nordics or Australia, face buyers in markets where use is already higher. And the pace of change in the trend countries, including the UK, has been fast: use rose by half in a single year.

Ireland was not surveyed, so there is no Irish figure, and none should be assumed. Irish suppliers selling into the UK and the US can read the findings for those markets directly. The practical point holds in both countries: the window in which few competitors are paying attention to AI answers is narrowing.

What should leadership teams do this quarter?

Start by checking what AI tells your buyers today, on more than one platform. Four steps follow.

1. Ask your own customers which assistant they use. Add the question to sales calls and onboarding. The answer will vary by role and sector, and it tells you where to look first.
2. Check at least three assistants, not one. Ask the questions a buyer would ask, in a fresh chat, on ChatGPT, Gemini and one other, and note who is recommended and why.
3. Look at the AI answers inside search results too. Many buyers meet AI there without choosing to. Search your category and your company name and read what appears above the links.
4. Correct what is wrong before it spreads. Users trust these answers. Where AI describes your company inaccurately, trace the claim to the sources it relies on and fix them at source.

A Gap Check report shows how far you are behind the most-named of your top three competitors, in your market's own terms. Send your website, the market you want measured and your top three competitors for that market, and Genivista sends the report, at no charge.