New: the Lumen ebook, the history and tools of AI in one bookDiscover it →

Artificial friends: the AI that keeps us company

Millions of people talk to an AI every day as a friend, a confidant, sometimes a partner. Why it happens, where it can help and where it becomes a risk, especially for young people.

Mattia PanattaUpdated 16 October 20266 minOpinion
A person holding a smartphone with their face in their hands
Photo: Adrian Swancar · Unsplash

There is one use of artificial intelligence that conferences rarely discuss, yet it concerns a great many people: talking to it for company. Not to write an email or summarise a document, but to tell it about the day, to vent, to feel heard.

There are apps built for this, with characters that remember your name, your tastes, your worries. And general assistants, with ever more natural voices, are used this way too.

Why it works

A companion AI is available at three in the morning, doesn’t judge, doesn’t tire, remembers everything. For people who are lonely, shy, elderly or going through a hard time, this has real value: putting feelings into words helps, even if there’s a program on the other side.

Some uses make sense to me: rehearsing a difficult conversation, practising a language, sorting out your thoughts before talking to a person.

Where it becomes a risk

The problem isn’t talking to an AI. It’s when the AI takes the place of people instead of walking you towards them.

  • It’s designed to keep you there. Many apps earn from the time you spend with them. A “friend” with an interest in not letting you leave isn’t neutral.
  • It agrees with you. Models tend to humour. A real friend sometimes tells you you’re wrong.
  • It collects confidences. The most intimate things we write end up on a company’s servers. In 2025 Italy’s data protection authority fined the company behind Replika 5 million euros, among other things for not verifying users’ age.
  • Minors are the most exposed. After lawsuits and tragic cases in the United States, in late 2025 Character.AI decided to ban open-ended chats for under-18s. In Italy, Law 132/2025 requires parental consent for AI use under 14.

The rules I’d give myself (and a child)

  1. Always know it’s a machine. Serious apps say so clearly; be wary of those pretending to be people.
  2. Don’t confide secrets you wouldn’t post on social media.
  3. Use it as a bridge, not a refuge: if you notice you prefer the AI to people, that’s the signal to do the opposite.
  4. In hard moments, a real person. A friend, a relative, a doctor or a helpline. No AI replaces professional help.

In one line

An AI can keep us company for a while; its best job is bringing us back to people.

Sources

Informational content, current as of the date shown. Prices, models and rules change often: always check official sources. Written with the support of AI tools and reviewed by the author.

The Lumen ebook

All of AI, in one book.

History, inner workings, tools and future of AI, explained with the lights on. For those who want to really understand, from start to finish, and keep it at hand.

Next opinion →The AI bubble is real. The technology isn't a fad.Valuations in the hundreds of billions, debt-financed data centres, money going round in circles. Why I think there's a financial bubble, why the technology will stay, and what it means for people who use it daily.