They hear, but do they care? How AI chatbots are reshaping the art of listening

Artificial intelligence may lack emotions, but for a growing number of people, it is proving to be a better listener than humans.

“I know it’s a machine, but it listens when I need it to,” says Anna, a Ukrainian woman living in London who regularly uses the paid version of ChatGPT. The BBC has changed her name to protect her identity.

What Anna values most is not the chatbot’s advice, but its silence — the absence of interruption, judgement or emotional projection. Unlike friends and family, who often rush to conclusions, the AI gives her space to reflect.

“I have a history with it,” she explains. “It understands my issues and responds in a way that works for me.”

That difference became especially clear after the end of her recent relationship. While loved ones quickly dismissed her ex-partner, the chatbot allowed her to explore conflicting emotions without pressure or criticism. The lack of judgement, she says, helped her understand herself better.

Anna’s experience is increasingly common. Research published by the Harvard Business Review shows that in 2025, therapy and companionship emerged as the most frequent use of generative AI tools such as ChatGPT — ahead of productivity, research and entertainment.

Even more striking are findings suggesting that AI responses are now often rated as more compassionate than those written by humans, including trained crisis hotline workers. Researchers say this is not because machines possess empathy, but because they consistently offer what many people rarely receive: uninterrupted, non-defensive listening.

In studies where participants were later told whether responses came from humans or AI, chatbot-generated replies were still judged to be more understanding, validating and caring. Other research found that users reported feeling less distressed, more hopeful and more emotionally supported after engaging with AI-generated text than with human responses.

The appeal lies in predictability. AI does not interrupt. It does not personalise the story or offer unsolicited solutions. It mirrors emotions, reflects language and allows users to take the lead.

Experts caution, however, that this is simulated empathy, not real understanding. AI systems draw from vast datasets of human interaction, learning how compassion is expressed — not how it is felt.

Yet the popularity of AI listeners raises an uncomfortable question: if machines are outperforming people at listening, what does that say about human communication?

The rise of conversational AI may not be teaching us how to replace human connection — but it is exposing how rarely we truly listen.