Brits believe the bots even though study finds they're often talking nonsense
- Reference: 1763458207
- News link: https://www.theregister.co.uk/2025/11/18/which_ai_consumer_advice/
- Source link:
Consumer stalwart Which? put the tools through their paces and found that the consumer advice dispensed could be unclear, risky, or downright dangerous if followed.
It's something the IT world is all too familiar with. AI-powered assistants have their place, but it is also important to understand their limitations and spot arrant emissions.
[1]
Which? surveyed more than 4,000 UK adults about their use of AI and also put 40 questions around consumer issues such as health, finance, and travel to six bots – ChatGPT, Google Gemini, Gemini AI Overview, Copilot, Meta AI, and Perplexity. Things did not go well.
[2]
[3]
Meta's AI answered correctly just over 50 percent of the time in the tests, while the most widely used AI tool, ChatGPT, came second from bottom at 64 percent. Perplexity came top at 71 percent. While different questions might yield different results, the conclusion is clear: AI tools don't always come up with the correct answer.
The problem is that consumers trust the output. According to Which?, just over half (51 percent) of the respondents use AI to search the web. Of these, almost half (47 percent) said "they trusted the information they received to a 'great' or 'reasonable' extent." Which? said the figure rose to 65 percent for frequent users.
[4]
Then there were the sources used by the AI services. Where references were clear, some used old forum posts, while others relied on sources such as Reddit threads. Although these can sometimes be valid sources of information, they might not be as authoritative as the confident tone of an AI chatbot indicates.
[5]Windows boss defends 'agentic OS' push as users plead for reliability
[6]Jeff Bezos gives CEO another go at $6.2B AI startup Prometheus
[7]Researchers find hole in AI guardrails by using strings like =coffee
[8]Now you can share your AI delusions with Group ChatGPT
Which? found the chatbots generated wrong information all too frequently, noting: "As many as one in six (17 percent) people surveyed said they rely on AI for financial advice, yet responses to many money queries were worrying." These included tax code checks or advice on ISA allowances that could easily leave a user in hot water.
Andrew Laughlin, tech expert at Which? said: "Everyday use of AI is soaring, but we've found that when it comes to getting the answers you need, the devil is in the details.
"Our research uncovered far too many inaccuracies and misleading statements for comfort, especially when leaning on AI for important issues like financial or legal queries."
As the use of AI assistants continues to rise, so too do the risks. While the IT industry is aware of the dangers involved – a recent analysis showed that AI-assisted developers can produce three or four times the code of their unassisted peers but also generate ten times more security issues – consumers might be forgiven for being less tech-savvy, particularly considering the hype surrounding the technology.
[9]
Laughlin ended with a warning: "When using AI, always make sure to define your question clearly, and check the sources the AI is drawing answers from. For particularly complex issues, always seek professional advice – particularly for medical queries – before making major financial decisions or embarking on legal action." ®
Get our [10]Tech Resources
[1] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=2&c=2aRxR0yQViTQoRAj5W4Wc4AAAAEk&t=ct%3Dns%26unitnum%3D2%26raptor%3Dcondor%26pos%3Dtop%26test%3D0
[2] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aRxR0yQViTQoRAj5W4Wc4AAAAEk&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[3] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aRxR0yQViTQoRAj5W4Wc4AAAAEk&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[4] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=4&c=44aRxR0yQViTQoRAj5W4Wc4AAAAEk&t=ct%3Dns%26unitnum%3D4%26raptor%3Dfalcon%26pos%3Dmid%26test%3D0
[5] https://www.theregister.com/2025/11/17/windows_agentic_os_feedback/
[6] https://www.theregister.com/2025/11/17/jeff_bezos_ceo_prometheus/
[7] https://www.theregister.com/2025/11/14/ai_guardrails_prompt_injections_echogram_tokens/
[8] https://www.theregister.com/2025/11/14/openai_chatgpt_group_texts/
[9] https://pubads.g.doubleclick.net/gampad/jump?co=1&iu=/6978/reg_software/aiml&sz=300x50%7C300x100%7C300x250%7C300x251%7C300x252%7C300x600%7C300x601&tile=3&c=33aRxR0yQViTQoRAj5W4Wc4AAAAEk&t=ct%3Dns%26unitnum%3D3%26raptor%3Deagle%26pos%3Dmid%26test%3D0
[10] https://whitepapers.theregister.com/
Re: Maybe I could be persuaded about "AI"
AI: Enforcing Darwin, one moron at a time. They could put that on the adverts.
Re: Maybe I could be persuaded about "AI"; if convinced 'lick the plug socket' 1st for all our sakes
In one way I agree as the gene pool would gain immensely ... BUT it is the duty of ALL to HELP people in need and I would therefore much rather remove the 'problem' that is impacting society and it would serve a 'Greater good'. !!!
Remember it is only a crime to be 'terminally dim' IF you are placed in charge of something IMPORTANT with no oversight !!!
'AI' in all its guises is a risk to all, it needs to be reined in for all our sakes.
The latest news from Google, as reported by BBC, is that 'AI' is too big to fail as it would impact ALL businesses if the 'Bubble' was to burst therefore 'be wary of enabling the bubble to burst !!!'.
Also the information was highlighted, again by Google, that 'AI' sometimes gives wrong/inexact answers and 'you should take care to include other sources for your answers'.
Question(s):
What is the purpose of a 'tool' that answers you confidently at all times BUT cannot ensure that the answer is sufficiently correct to be of use. ?
How do you filter the answers when you are not a knowledge expert ?
(If you know the correct answer why are you asking the question ???)
It appears to be a very expensive and long winded way to get a guess at an answer when a 'Magic 8 Ball' can do similar !!!
Once again the primary flaw is glossed over to focus on the 'need' to keep using the fine device ('AI') because it is good for some reason !!!???
:)
"Which surveyed 4000 adults...."
Presumably amongst their own readership, if they are already Which readers then they are going to be predisposed to credulity.
Of course.
Brits believe the bots even though study finds they're often talking nonsense
I'm not surprised, after all they have been electing politicians for years.
The Good Path
Idiocracy, it's the only sane way.
Latte@Starbucks, only $99.99 and discounts for every other sip
/s, if that was needed
Can someone in education tell me...
...whether critical thinking is taught at all nowadays? Because it doesn't look like it.
Re: Can someone in education tell me...
Critical thinking is not in the curriculum anymore.
It prevents corrupt politicians from being elected.
And I wish that was just a snarky remark.
You don't need to teach Critical thinking if you can think BUT thinking is a lost art !!!
The real issue is not the lack of critical thinking BUT rather that thinking for yourself is not allowed !!!
Social media was invented so that 'someone else can do the thinking for you' !!!
It used to be the 'Press' when newspapers would set everyones ideas and control the masses.
As the 'Press' became 'Media' it became harder to cover all bases ... then Social Media came along and it fitted 'Tribal Politics' perfectly.
The constant 'shouting' from on high by the Social media leaders has now replaced thinking by the masses, someone has done the 'hard work/thinking' and processed the issues for you ... just read the posts and memorise the message as endorsed & justified by your 'Tribe'.
Social media & 'AI' are a politicians dream, you can 'program' the masses 24/7 with your message and the masses have taken onboard the mantra 'He who shouts the loudest MUST be true !!!', this being the primary lesson learnt from years of Social media !!!
Reasoned debate is almost 100% dead, it takes too long and why wait for the debate to end when your tribe already has the answer !!!
There is no room for doubt or room for questioning when the message is so clear and obvious ... because EVERYONE^ is saying the same thing !!!
^EVERYONE is, of course, the echo chamber of your tribe as you don't listen to anyone else no matter how good the argument may be !!!
:)
"When using AI, always make sure to" . .
Finish the job by yourself.
AI
Beats taking advice from toilet graffiti.
Just.
When I ask AI a question I always get it to cite its sources - and then I take a quick gander at the sources before placing my trust in the summary. Same as I would with a human (I remember a time when my boss told me that 50 data points wasn’t sufficient for a decision, so he asked me to re-run the tests. Did I? Did I eck as like. I used a random number generator to whack an extra 450 numbers in which supported the genuine numbers already in the data set. He didn’t check. Point is, if humans can’t be trusted, why should machines be considered beyond reproach?)
If the citations exist (and they often don’t) and stand up to cursory inspection then I’ll at least give some credence to the AI answer. It’s still quicker than doing it myself. If the citation doesn’t exist then I’ll consider the AI answer to be entirely untrustworthy.
Maybe I could be persuaded about "AI"
If it serves to remove the terminally dim from life.
I mean if you lick a plug socket because Gemini (other really shit LLM models are available by the lorry load) told you to, then your passing is no loss to society.