AI for Seniors: A Friendly Starter Guide to ChatGPT, Gemini and Privacy

ChatGPT and Gemini are not difficult. They are, however, easy to misunderstand, and most of the misunderstanding costs people either money or privacy. So let us deal with both, in the order in which they usually come up.

What ChatGPT and Gemini actually are

Both are chatbots: programs that generate text, images and speech in response to whatever you type or say. Neither is a search engine, a database, or an authority on anything.

They work by predicting which words are likely to follow yours. That is why they so often sound confident and are sometimes simply wrong — a habit worth remembering before you act on an answer about medication, taxes or a contract.

Both are available across the European Union, including Czechia, Slovakia, Poland, Germany, France, Spain and Italy, and both handle European languages well. Czech and Slovak are supported, though the smaller the language, the more often you will want to double-check a result. Both have free tiers that are genuinely usable. You do not have to pay anything to find out whether this is for you.

Which one should you start with?

Short answer: if you already use Gmail or an Android phone, Gemini is the path of least resistance, because it lives inside the account you already have. If you want the tool with the most guides and the largest community of users to ask for help, start with ChatGPT. The differences matter far less than the habit of using one of them regularly. If you prefer a structured walkthrough to an article, our AI courses cover the basics in plain language.

Your first ten minutes, step by step

1. Install only the official app. Go to the company's own website and follow the link to the App Store or Google Play from there. App stores are full of look-alike apps that charge for something that is free, or worse.
2. Sign up with an email address you actually control, and use a long password. Then turn on two-step verification — a code sent to your phone. It takes two minutes and it is the single most useful security setting you will ever change.
3. Ask one real question. Not "hello". Something you actually want to know: "Explain in plain language what a deductible means in home insurance." The tool answers to the level of detail you ask for, so ask for what you need.
4. Press the microphone. Most people over sixty type more slowly than they speak, and the voice button removes that friction entirely.
5. Set the privacy switches described further down. Do this on day one, while you still remember your password.

Six things people actually use it for

Letters and forms. "Write a polite email to my energy supplier asking them to explain a €60 increase on my last bill. Keep it under 150 words and keep the tone calm." You copy, adjust, send. This alone justifies the effort of learning to use it.

Understanding paperwork. You can photograph a letter and ask what it means. Be careful: photograph only what is necessary, and black out ID numbers, account numbers and anything medical that identifies you.

Cooking and shopping. "I have potatoes, eggs and spinach. What can I make for two people, in under 40 minutes?"

Learning and hobbies. Language practice, gardening questions, help with a watercolour composition, or turning a family story into readable prose.

Travel. "Explain how to buy a tram ticket in Lisbon, step by step, for someone who doesn't want to use an app."

Health questions — with care. Use it to understand terminology and to prepare questions for your doctor. Never use it instead of your doctor. It cannot examine you, and it cannot be held responsible.

Talking instead of typing

Voice mode is the biggest usability win for anyone with arthritis, poor eyesight, or a lifelong dislike of keyboards. You press the microphone, speak normally, and it answers out loud. You can interrupt it mid-sentence, and it will follow you.

Two things to know. First, voice conversations are stored much like typed ones, so the same privacy rules apply. Second, voices can be copied. Scam calls using a cloned voice of a relative have been reported across Europe. If someone rings asking for money in a voice you recognise, hang up and call that person back on a number you already have.

What it costs in Europe

Most people never need to pay. The free tiers are enough for letters, questions and a few photos a day. If you do hit the limits, here is roughly what the upgrades cost — European prices normally include VAT, while US list prices do not, which is why the numbers below look closer than the exchange rate suggests.

The Google option is quietly interesting if you already pay Google for extra storage: the subscription includes it, so the effective price can be lower than it looks. Prices change, so check the vendors' own pages before you subscribe — and note that both companies offer an annual discount in some markets.

Privacy: the switches worth flipping today

This is the part that matters most for European users, and the part almost nobody does.

In ChatGPT: open Settings, then Data controls, and switch off "Improve the model for everyone". Until you do, your new chats may be used to train future models. There is also a Temporary chat mode — a conversation that is not saved to your history and not used for training, though it is kept briefly for safety reasons. For a formal request to see or delete everything OpenAI holds about you, use their privacy portal; under the GDPR you have the right to access, correction, deletion and a portable copy of your data.

In Gemini: open your Google Account, go to Data & privacy, and turn off Gemini Apps Activity. Google states that with it off, conversations are not saved to your account and are kept for up to 72 hours. You can also review and delete past activity at any time.

Three rules that survive any setting: do not type in passwords, bank details or ID numbers; do not paste in other people's personal data — if you run a club or a small business, the GDPR applies to you too; and if you would not write it on a postcard, do not type it into a chatbot.

Europe does enforce this. In December 2024, Italy's data protection authority fined OpenAI €15 million over the ChatGPT data breach and transparency failures — a reminder that these services operate under European rules, not outside them.

What the EU AI Act means in practice

The AI Act entered into force on 1 August 2024. Since 2 August 2025, providers of general-purpose AI models — which is exactly what ChatGPT and Gemini are — have had formal EU obligations: technical documentation, a copyright policy, and a public summary of the training data.

From 2 August 2026, transparency rules kick in, including the requirement that you be clearly told when you are interacting with an AI rather than a person. For you as an individual, nothing about your daily typing changes. The practical meaning is simple: the companies now have a regulator to answer to, and the GDPR already protects your data today.

When not to trust it

It is weakest where the stakes are highest. Do not rely on it for medical decisions, legal advice, tax calculations, or anything local and current — opening hours, bus times, today's prices. It will sometimes invent a perfectly plausible answer rather than admit it does not know. It also tends to agree with you, which feels pleasant and is not always useful.

One habit solves most of this: ask "how do you know that?" and check anything important on an official website. Use it to prepare your questions for a professional, not to replace the professional.

Start with one genuine question this week — a letter you have been putting off, a term you never understood. That is enough. Everything else follows from realising that it is a tool, not an authority, and that you are allowed to tell it when it is wrong.