What should you tell older relatives about how to use AI?

AI chatbots can undoubtedly make everyday life a lot easier for older people. However, their use also raises many questions and concerns.

If you’ve been your family’s unofficial tech support for the last 20 years or more, then it’s increasingly likely that you’re also going to be the person dragooned into answering questions about AI and chatbots from your parents and grandparents, alongside troubleshooting their dodgy WiFi signal and helping them find that long-lost file.

And increasingly, older generations are asking questions about AI. Research published by the Pew Research Center suggests 23% of Americans aged 65 and over have used an AI chatbot, although just 6% feel extremely or very confident using one.

So older generations are becoming AI-curious, but there’s also a wealth of issues that could go wrong. So, how do we actually ensure that everyone can harness the opportunities AI offers, without being scared off by the doom and gloom that surrounds it?

Get over the hump

The first thing to make clear to older relatives is how AI actually works and what its strengths and limitations are. Chatbots like ChatGPT, Claude, and Gemini are designed to respond to questions and instructions in natural language.

You don’t need to learn complicated commands or understand how the technology works to make them work.

One of the best ways to demonstrate that is to get a relative to do something practical. That could be something small and simple for them to show off the skills that AI has. Maybe ask a chatbot to suggest a recipe using ingredients already in their fridge, to help plan a holiday, or to explain an unfamiliar technical term.

Unlike traditional devices, AI allows and encourages experimentation. The conversational aspect and the way you can interact with an AI model through chatbot interfaces allow people to ask questions in their own words, which can be very empowering for older relatives.

But there’s an important catch – and one that can sometimes hoodwink those less experienced with tech.

Be honest, even if AI won’t be

One of the most important things to explain to relatives is that AI chatbots can be confidently wrong. They hallucinate: it occurs when AI generates information that sounds plausible but is either incorrect or entirely invented.

Tell your relatives that what they read from an AI chatbot might be made up, a historical fact could be wrong, an incorrect website address could be shown, or the chatbot could just misunderstand their question.

As a result, media literacy is important for older relatives, just as it is for younger ones. Tell older relatives they should use AI as a starting point for research, rather than the final authority. That’s particularly important for anything to do with medicine, finance, or family planning.

Sharing sensibly

Faced with the opportunities that AI offers, it can be very easy for older relatives to simply give over everything they know and everything they own to a chatbot for convenience. But remind them that they shouldn’t enter passwords, banking information, or sensitive medical details without understanding how that information will be handled.

Most major services offer privacy controls. ChatGPT allows users to switch off the use of their conversations for model training through its settings – so set that up for them and show them where it is.

It’s also worth reminding them that any technology they use can be used by bad actors, too. AI-powered scams are becoming a real concern. Criminals can use AI to imitate someone’s voice, to create convincing phone calls, and then – perhaps – to demand money.

So it’s always worth asking your relatives to have a healthy skepticism about whatever technology they’re using – and that’s particularly the case with AI.