We have hundreds of connections. So why are we confiding in machines?

By Tahani Elghazaly

 For all the ways technology has made us more connected, many people still struggle with one very basic thing: finding someone they trust enough to talk to when life gets hard.

That tension sits at the centre of a recent Canadian report on young people using generative AI for mental health-related support.

A Jack.org survey of 395 young people in its network found that one in three respondents used generative AI at least once a month for mental health support. Nearly half of those users had not told anyone they were doing it.

The obvious question is why.

Part of the appeal is practical. AI is available at any hour. It does not appear impatient. It does not interrupt. It does not react with embarrassment or judgment when someone admits to feeling anxious, lonely or overwhelmed.

For someone who is unsure how another person might respond, that can feel easier than starting a difficult conversation.

But ease is not the same as connection.

Statistics Canada has found that young people who feel less connected to their communities are much more likely to report regular loneliness. That matters because loneliness is rarely just about being physically alone. People can be surrounded by classmates, co-workers, relatives and online contacts and still feel that there is no one they can speak to honestly.

This is where the conversation about AI becomes more interesting.

The technology may be filling a gap that already existed.

It can help users put feelings into words, organize thoughts or rehearse a conversation they are nervous about having. Those functions may be useful. But they also reveal something uncomfortable: some people may find it easier to type their fears into a chatbot than say them out loud to someone they know.

That should concern us less as a technology problem than as a relationship problem.

Mental health experts have cautioned against treating generative AI as a substitute for professional care. A chatbot cannot assess risk the way a trained clinician can. It cannot notice the change in a friend’s behaviour over several weeks. It cannot decide to show up at someone’s door because something feels wrong.

And yet simply telling young people to stop using AI would miss the point.

If a person is choosing a machine over another human being, the more useful question is not whether the technology is good or bad. It is whether the people around them feel safe enough to approach.

That is where families, schools, workplaces and communities have more influence than they may realize.

Often, support starts with something very small: asking a question and leaving enough silence for an honest answer.

We are used to asking, “How are you?” as a formality. The question often expects a short reply and a quick transition to something else.

Real listening requires more patience than that.

It also requires resisting the instinct to solve everything immediately. People do not always need advice the moment they begin speaking. Sometimes they need room to explain what is happening without being corrected, compared or rushed.

Research on social connection continues to show how strongly belonging is tied to well-being. The lesson is not that every difficult moment requires therapy or that every conversation must become serious.

It is simpler than that.

People need to know that there are relationships in their lives where honesty does not come with a penalty.

AI may remain part of how younger generations think, work and even process emotions. That is unlikely to change.

The real challenge is making sure technology does not become the only place where they feel heard.

A chatbot can respond instantly.

A person can do something more important: notice, care and stay.