AI Netiquette · AI Use Best Practices

What is AI, in simple words?

Here, “AI” mainly means tools that can generate text, images, code, or audio from your requests (prompts).

They are not magic, and they are not people. They are programs trained on huge amounts of data to guess “the next best word (or pixel)”.

Different AI tools are good at different things. Some are like chatty assistants, some are specialised (for translation, coding, images, audio), and some are hidden “inside” apps you already use.

Important to remember

AI can be very useful, very fast.
AI can also be wrong, biased, or outdated.
You are always responsible for what you do with it.

What you should do

Write clear prompts. Explain what you need, give context, and ask for examples.
Double‑check important answers. For health, money, law, or work decisions, always verify with trusted human sources.
Protect your privacy. Remove names, addresses, passwords, and confidential data before sending text to an AI tool.
Use AI as an assistant, not a boss. Let it help you think, summarise, draft, and translate.
Tell the AI who you are. Saying “I’m a beginner”, “I’m 15”, or “I’m a lawyer” helps it adapt the explanation to your level.
Ask it to show its work. For complex topics, ask for steps, sources, or alternative options instead of just a final answer.

Good vs bad prompt

❌ Bad

“Do my homework.”

✅ Good

“Explain photosynthesis like I am 12, using a short paragraph and a simple analogy, then give me 3 quiz questions I can answer.”

✅ Also good (work email)

“Draft a polite email to a colleague to ask for an update on project X. I want it to sound friendly but professional. Then give me 2 shorter alternative versions.”

What you should not do

Do not paste secrets. No passwords, ID numbers, medical records, internal company documents, or client data.
Do not blindly trust. AI can “hallucinate” – invent facts, quotes, or sources.
Do not use AI to cheat. Copy‑paste work for exams, theses, or job tasks can be plagiarism and may be detected.
Do not use AI to harm. No scams, hate, discrimination, or harassment. Laws still apply.
Do not overshare about others. Avoid sending names and details about friends, family, patients, or clients unless you are 100% sure it’s allowed.

Red‑flag uses

“Write a convincing email to steal someone’s password.”
“Help me bypass this exam system.”
“Generate insults for this group of people.”

Many AI tools block these requests. Even if they don’t, you are still responsible for the consequences.

Which AI tools to use, and where?

Main types of tools

Chat assistants. General‑purpose tools where you type or speak with an AI (like this one). Good for explanations, drafts, ideas, and coding help.
Specialised apps. Translation, grammar checking, slide generation, image creation, music, video, coding copilots, and more.
AI inside other software. Email suggestions, autocomplete, AI features in office suites, design tools, and development tools.

Where AI can help the most

Learning & study. Ask for explanations, analogies, summaries, and practice questions.
Everyday writing. Emails, messages, letters, CVs, cover letters, social posts.
Understanding documents. Summarise long texts, highlight main points, suggest questions to ask a human expert.
Brainstorming. Generate lists of ideas, titles, angles, or examples – then you choose what is useful.

For medical, legal, or financial decisions, treat AI as a pre‑reading tool, not as the final authority.

Risks & Benefits

Benefits

Faster drafts, summaries, and translations.
Support for people with reading or writing difficulties.
New ideas for creativity, study, and work.
24/7 availability: the AI does not get tired or impatient.

Risks

Wrong or misleading information.
Biased or offensive outputs.
Leakage of private or confidential data.
Over‑reliance: “I can’t think without AI anymore.”

Rule of thumb

The more important the decision (health, money, job, legal matters), the more you should:

Use AI only to understand or prepare.
Verify with trusted, human experts.
Keep the final decision human.

Glossary – AI words explained simply

AI (Artificial Intelligence)

Computer systems that can perform tasks that normally need human intelligence, like understanding language, recognising images, or making suggestions.

Model

The “brain” of the AI: the mathematical system that has learned from lots of data and now produces answers or content.

Prompt

What you type or say to the AI to tell it what you want. A good prompt is specific, clear, and includes context.

Hallucination

When the AI confidently gives an answer that sounds correct but is actually wrong or invented (for example, a fake quote or source).

Bias

Systematic unfairness in the AI’s outputs. This can happen when the data the model learned from reflects real‑world stereotypes or imbalances.

Context / history

The previous messages in your conversation. Many tools use this to keep track of what you are talking about, but they can forget older parts.

FAQ for non‑experts

Does AI “know” everything?

No. It generates answers based on patterns in the data it was trained on. It does not “understand” like a human and can be wrong or outdated.

Is AI spying on me?

Many services store your prompts to improve the system. Avoid sending information you wouldn’t be comfortable sharing with a stranger.

Can I use AI for school?

Often yes – as a tutor: to explain, summarise, or quiz you. Using it to hand in work as if it were yours may break school rules.

Can I use AI for work?

Many companies allow AI for drafts and ideas, but forbid sharing internal or client data. Always check your workplace policy.

Can I talk to AI in my dialect?

Many tools understand several languages and some dialects. If it struggles, try mixing in standard language or ask the AI to “translate this dialect into standard Italian/English first”.

What if the answer feels “off”?

Trust your common sense. You can ask the AI to re‑check, to show sources, or to give alternative views – and you can always ignore an answer that does not convince you.