They Hear You—But Are They Really Listening? What AI Reveals About Better Communication

“I am aware it’s a machine but it’s super convenient and knows how to listen well whenever I need it,” says Anna, a Ukrainian living in London. She is talking about her regular use of the premium version of ChatGPT, a chatbot powered by artificial intelligence.

What Anna – the BBC is not using her real name to protect her identity – finds particularly valuable isn’t necessarily the AI’s advice, but its ability to give her space for self-reflection.

“I have a history with it, so I can rely on it to always understand my issues and communicate with me in a way that suits me,” she says. She is aware that this might seem odd to many people, including her friends and family, which is why she has asked to remain anonymous.

But when she recently broke up with her boyfriend she found the AI’s patient listening offered something that her protective friends and family couldn’t provide with their immediate judgments about her ex-partner – “he’s an idiot”.

Instead, the absence of judgement created an opportunity for self-understanding as she unpacked her mixed emotions.

And Anna is not alone: recent Harvard Business Review research shows that in 2025 therapy and companionship was the single most common use of generative AI through the family of tools like ChatGPT, which can carry out a conversation much like a person.

Strikingly, studies show AI-generated text responses are now rated as more compassionate than those written by humans – even when those humans are trained responders from crisis hotlines. This isn’t because AI is genuinely more compassionate, but rather a sobering indictment of how rarely we listen in a non-judgmental way.

When researchers disclosed the identity of the response authors, evaluators still judged responses from ChatGPT responses to be more understanding, validating and caring – revealing how hungry people are for uninterrupted, non-defensive listening. In another study, people reported experiencing more hope, less distress and less discomfort after interacting with AI-generated responses compared to humans.

It is worth remembering that these AI chatbots are not displaying real empathy, but rather simulating it based on what they have learned from huge datasets of human interactions.

The irony that an algorithm powered by a large language model – the type of machine learning that underpins many AI chatbots – might be perceived as a better listener than an authentic human reveals important insights about our human listening shortcomings. It’s when our agendas, backstories and emotional triggers run the show, that true deep listening becomes thwarted.

None of this is to suggest we should trade real person relationships for large language models. But it does suggest there are some lessons that we humans can learn from these code-based listeners.