Most Britons Say Please to AI, a Survey Finds
A fonio.ai survey claims about seven in ten Britons are polite to AI assistants, with some even apologizing when exchanges go wrong.
According to TechRadar, a survey by the AI company fonio.ai claims that roughly 70 percent of people in Britain say "please" or "thank you" when they speak to AI assistants. The outlet presents the findings as evidence that people are becoming more comfortable with the tools and are extending courtesy beyond basic greetings.
The same survey, as TechRadar describes it, found a growing number of respondents ask an AI assistant how it is doing, a habit usually reserved for conversations with other people. Some respondents also said they apologize to the assistant when something goes wrong in an exchange.
TechRadar places the results against a backdrop of public unease about AI. The outlet notes that worries about the technology taking over have continued to rise even as everyday use spreads, and it frames the politeness findings with a light, slightly humorous tone.
The reporting does not say how the survey sample was selected, how many people took part, or how the questions were worded. Because the results are based on what people say they do, they describe reported behavior rather than a direct record of conversations with AI tools.
Background
Conversational AI has moved quickly from novelty to routine. Voice assistants on phones and smart speakers have been part of daily life for more than a decade, and chatbots built on large language models have pushed the experience further by producing long, natural-sounding replies. Many people now use these systems for email drafts, homework help, travel planning and quick questions, often in a typed chat window that resembles a messaging app.
Psychologists and computer scientists have studied for years how people treat machines that talk back. A well-known line of research from the 1990s found that people tend to apply social rules, such as politeness and reciprocity, to computers even when they know the machine is not a person. That tendency helps explain why many users say thank you to a device, or feel a flicker of embarrassment when they are curt with one.
Politeness toward AI also has a practical dimension. Some users say courteous phrasing gets better results, though the evidence on that point is mixed and depends on the task and the system. Others simply prefer to keep a habit of civility, reasoning that the way they communicate with a tool may shape how they communicate with people.
Surveys about consumer attitudes toward technology are common, and companies that sell AI products often publish them as marketing research. Such studies can be useful for spotting trends, but they usually rely on panels recruited by the sponsor, use questions written to produce headline-friendly results, and are rarely subject to independent review. Readers generally get the most value from them when they are compared with academic work and with data from neutral organizations.
Why it matters
Whether people are polite to AI may look like a minor cultural curiosity, but it touches on larger questions about trust and design. When people anthropomorphize a tool, they may extend it more confidence than its accuracy warrants, or they may feel reluctant to question its answers. Apologizing to an assistant after a misunderstanding suggests users see the exchange as a relationship, which can affect how they judge errors and whether they check the output.
For companies, the findings point to a design challenge. Developers have to decide how much personality an assistant should show, whether it should mirror a user's tone, and how it should respond to rudeness or to users who treat it as a confidant. Those choices shape the emotional tone of the product and can influence how much people rely on it for advice.
The story also highlights the gap between how people describe their habits and how they behave. Self-reported politeness is easy to measure in a survey, but actual patterns of use would require data that most providers do not share publicly. Without that, the most striking figures remain hard to verify.
For everyday users, the takeaway is less about manners than about attention. Courtesy is harmless, but it should not be mistaken for evidence that an assistant understands or cares. Keeping a clear sense of what these tools are, and checking what they produce, matters more than whether the request began with "please."
What to watch
The first question is whether the survey's methodology is published in full. Details about the sample, the questions and the way results were weighted would show how much confidence the figures deserve, and whether they can be compared with other research.
Independent studies will be the real test. Academic researchers and neutral polling organizations could examine whether politeness is widespread across age groups and regions, and whether it changes over time as people grow more familiar with AI tools.
Product design is another signal to follow. If major AI providers adjust how their assistants respond to courtesy, apologies or personal questions, that would show how seriously they take the emotional side of user behavior.
Finally, watch for any research that examines whether the way people phrase a request affects the quality of the answer. That evidence would give the politeness debate a practical footing beyond opinion and survey headlines.
Viewpoints
Commentary from our columns: three angles on this story.
The Skeptic open questions and weak points
A seven-in-ten figure is the kind of number that travels fast, and it deserves a closer look before anyone builds a theory around it. The survey comes from fonio.ai, a company with a commercial stake in how people feel about AI assistants. That does not make the findings false, but it does mean readers should ask who designed the questions, who answered them, and why the headline emphasizes friendliness rather than, say, how often people double-check what a chatbot tells them.
Self-reported behavior is also slippery. People tend to describe themselves as courteous, and they may answer a survey the way they think a thoughtful person should. Saying "thank you" to a chatbot after a useful answer is one thing; reporting that you apologize to software is another, and it is hard to know how many people do that in practice.
The bigger weak point is the leap from manners to meaning. Politeness toward a tool tells us something about habit and comfort, but it says little about whether people understand what the tool is doing. The most interesting question the survey raises is not whether Britons are nice to their assistants. It is whether that niceness is making them trust those assistants more than they should.
The User what changes for people who use it
For most people, the change here is small and personal. You open a chat window, ask for a recipe or help with an awkward email, and somewhere along the way you type "thanks" because it feels natural. Nobody needs a survey to explain that. What the research adds is a sense that many of us already treat these tools as something between a search box and a colleague.
That matters because the way an assistant feels can shape how we use it. If a chatbot seems friendly and patient, we may lean on it for things we would normally check elsewhere. A polite exchange can make a shaky answer sound more authoritative than it is. Keeping a simple habit of verifying important details, such as medical, legal or financial information, is more useful than worrying about whether you were rude.
It is also worth noticing how normal this has become. Apologizing to an assistant after a misunderstanding is an everyday reaction now, not a sign of confusion. Used with a clear head, that instinct can make the tools feel easier to work with. The practical advice is simple: be as courteous as you like, and keep your own judgment switched on.
The Optimist the hopeful reading
There is something quietly encouraging about people choosing courtesy when they talk to a machine. Politeness is a habit that reflects care for others, and the fact that so many people carry it into conversations with software suggests that the instinct to be decent has not been crowded out by new technology. Some might find that charming, and others might find it a little silly. Both reactions can be fine.
The hopeful reading also points to a possible design direction. If people naturally bring warmth to their assistants, developers have a reason to build tools that respond with clarity and patience, and that make it easy to ask questions without feeling judged. A well-designed assistant could reward good manners by being more helpful, and could nudge users toward honest, thoughtful exchanges rather than curt commands.
None of this needs to become sentimental. People are still the ones who decide how much to trust a tool and when to check its work. But the impulse to say thank you, or to ask how an assistant is doing, is a small sign that people want their technology to feel human-friendly. That is a reasonable thing to hope for as these tools become more common in everyday life.