Can AI Help Psychologists Show Empathy?

Key Points (For Readers on the Go)

• There is an audio version of this post available for people on the go.

• A new study suggests that AI-based practice may help people improve how they communicate empathy, not just how empathic they believe themselves to be.

• This has clear relevance for people in helping professions, especially psychologists, school psychologists, trainers, supervisors, and trainees.

• One of the most important takeaways is that feeling empathy and expressing empathy are not always the same thing.

• AI may have value here as a practice and coaching tool, not as a replacement for human supervision, clinical judgment, or real relationships.

• The findings are promising, but they come from a structured, text-based research setting, so they should be applied thoughtfully.

Introduction

There is an audio version of this post available for people on the go.

This post is based on the article Practicing with Language Models Cultivates Human Empathic Communication. It caught my attention because it explores a question that matters well beyond AI enthusiasm: can structured practice with a language model help people get better at showing empathy in words?

That question matters for many helping professions, but it is especially relevant for psychologists and school psychologists. Whether you are training graduate students, supervising interns, consulting with families, giving difficult feedback, or trying to strengthen your own communication, empathy is not just something you feel. It is something you have to convey.

That is where this study becomes especially interesting. It suggests that empathic communication may be more teachable, more observable, and more improvable than many of us assume.

Why This Study Matters for Psychologists and Other Helping Professionals

Psychologists are often expected to communicate care, understanding, and emotional attunement in difficult moments. The same is true for many other helping professionals. But in psychology especially, that expectation shows up constantly.

A school psychologist may need to explain difficult evaluation findings to a parent. A clinical psychologist may need to respond to grief, frustration, or shame. A supervisor wants to help trainees learn to convey empathy.

Many people assume that if they care deeply, that caring will naturally come across. But that is not always true. A person can feel concerned and still respond in ways that sound rushed, overly clinical, too advice-heavy, or emotionally flat.

This study is useful because it focuses on that gap. It is not mainly about whether people have empathy as a trait. It is about whether they can communicate empathy in a way another person would experience as supportive.

What the Study Did

The researchers created a platform in which participants engaged in text-based conversations with an AI system role-playing someone going through a difficult situation. These scenarios included emotionally challenging experiences such as grief, job loss, and family illness.

Participants practiced writing supportive responses. Some received no feedback. Some received brief instructional video content. Some received personalized AI feedback on their responses. Another group received both.

The study was large and structured, with hundreds of participants and thousands of messages. The researchers then evaluated participants’ communication using several empathy-related dimensions. These included things like validating emotions, encouraging the other person to say more, demonstrating understanding, and avoiding less helpful patterns such as dismissing emotions or jumping too quickly into advice.

That framework alone is useful. It reminds us that empathy is not just a vague feeling. It can be broken down into recognizable communication behaviors.

What the Study Found

The main finding was that personalized AI feedback helped people improve their empathic communication.

Participants who received AI-based coaching generally improved more than those who received no feedback. In many cases, they also improved more than those who received only brief instructional content. Importantly, the gains were not just about writing more or spending more time. The improvement was in the quality of the responses.

After receiving AI feedback, participants were more likely to do things that tend to support empathic communication, such as validating feelings, showing understanding, and inviting elaboration. They were also less likely to rely on patterns that often weaken empathic responding, such as unsolicited advice, self-focused replies, or dismissive statements.

For psychology trainers, that matters. It suggests that practice plus targeted feedback may help people refine how they respond in emotionally significant moments.

The Most Important Insight: Feeling Empathy Is Not the Same as Expressing It

This was the most important takeaway for me.

The study suggests that people often think they are communicating empathy more effectively than they actually are. In other words, intention and performance do not always match.

That is highly relevant to psychology training. A trainee may leave a conversation thinking, “I was warm and supportive,” while the other person may have experienced the response as minimizing, rushed, or overly solution-focused. The problem may not be lack of caring. It may be lack of skill in how that caring is expressed.

This matters because we often talk about empathy as though it is primarily internal. But in practice, much of it is relational and observable. The other person only has access to what we communicate.

That has implications for how we train psychologists. We should not assume that empathic intent automatically produces empathic communication. Sometimes people need explicit guidance, examples, rehearsal, and feedback.

What This Could Mean for Training

For trainers, supervisors, and practicing psychologists, the study points toward a practical possibility: AI may be useful as a low-stakes practice partner.

That does not mean AI can replace supervision. It does not mean it can model the full complexity of human relationship-building. And it certainly does not mean psychologists should outsource empathic communication to a machine.

What it may mean is that AI can help people rehearse difficult responses before they use them in real life.

For example, a school psychology trainee might practice how to respond to a parent who is upset about unexpected evaluation findings. A practicum student in clinical psychology might rehearse a response to a client describing grief or hopelessness. A supervisor might use a vignette to help a trainee notice when they move too quickly into fixing, explaining, or reassuring.

That kind of repetition may have real value. Many people get very little deliberate practice in empathic responding outside of real, high-stakes encounters. AI may offer a way to slow down, try again, and refine language before it matters most.

Practical Examples Section

Graduate training

Instructors could give students a vignette involving grief, frustration, fear, or disappointment and ask them to draft an empathic response. Students could then revise that response after receiving structured feedback and discuss what changed.

Supervision

Supervisors could ask trainees to rehearse how they would respond in common difficult situations, such as a tearful parent meeting, a student in distress, or a client reacting strongly to feedback. The goal would not be to script people rigidly, but to help them notice patterns in their communication.

Practicing psychologists

Even experienced psychologists may want to strengthen how they phrase difficult conversations. AI could be used as a rehearsal space for wording, especially when someone wants to sound both clear and caring.

Communication skill-building

Programs might also adopt a simple reflection checklist:

• Did I validate the person’s feelings?

• Did I show understanding?

• Did I invite them to say more?

• Did I avoid shifting too quickly into advice?

• Did I avoid centering myself?

That kind of framework makes empathy more teachable.

Considerations

The findings are interesting, but they should be interpreted carefully.

First, this was a text-based study in a controlled setting. Real empathic communication involves much more than words on a screen. Tone of voice, pacing, facial expression, silence, body language, and relational history all matter.

Second, the scenarios were structured and somewhat contrived. That is common in research, but it means we should be cautious about assuming the findings transfer perfectly to real-world school, clinical, or supervisory relationships.

Third, helping relationships are shaped by culture, context, trust, power, and history. Those factors are hard to capture in a short simulated exchange. So while the study tells us something useful, it does not tell us everything we need to know about empathy in actual practice.

Fourth, the takeaway here is not that AI is empathic. The more responsible interpretation is that AI may help humans practice communicating empathy more effectively. That is a very different claim.

Final Takeaways

• This study suggests that AI-based feedback may help people improve how they express empathy.

• That finding is especially relevant for psychologists, school psychologists, supervisors, and trainers.

• One of the clearest lessons is that caring about someone is not always the same as communicating care effectively.

• AI seems most promising here as a practice tool, not as a substitute for supervision, human connection, or professional judgment.

• For people in helping professions, this study is a reminder that empathic communication is not just a trait. It is also a skill that can be strengthened.

If your program, clinic, school, or organization is thinking about how to use AI in thoughtful, practical, and ethical ways, visit my website to learn more about training and consultation services: https://lockwoodconsulting.net/

AI Use Disclosure - Portions of this post were drafted with the assistance of an AI writing tool and revised by the author for accuracy, clarity, and professional judgment.