AI in Therapy: Should Parents Be Excited or Concerned?

Artificial Intelligence (AI) seems to be everywhere these days. It’s helping us write emails, answer questions and even organise our shopping lists. But what about therapy? Can AI really help children and families?

The short answer is yes—but with an important catch. AI should support therapists, not replace them.

As allied health professionals increasingly use technology to improve services, AI is becoming another tool that can help children access therapy earlier, receive more personalised support and spend more time focusing on what really matters. Here’s what parents need to know.

What is AI in therapeutic services?

Artificial Intelligence refers to computer systems that can analyse information, identify patterns and make predictions. In healthcare, AI is designed to assist professionals by processing large amounts of information quickly and accurately.

You may already be using AI without realising it. Many therapy apps, communication devices, scheduling systems and online screening tools already include AI in the background.

Importantly, AI does not replace the knowledge, empathy and clinical judgement of a therapist. Instead, it can help therapists work more efficiently and make better-informed decisions.

How can AI help children and families?

Earlier identification and screening

One of the biggest challenges for many families is the long wait for assessments.

Researchers are developing AI systems that can analyse videos, speech, language patterns and behaviour to help identify children who may benefit from further assessment. For example, studies have shown that AI can analyse short home videos to identify behavioural patterns associated with autism with promising levels of accuracy.

This doesn’t mean AI diagnoses a child. Rather, it can help identify children who may need a comprehensive assessment sooner, potentially reducing waiting times and helping families access support earlier.

More personalised therapy

Every child is different. AI has the potential to help therapists tailor intervention to each child’s strengths, interests and goals.

By identifying patterns across therapy sessions, AI may help clinicians notice subtle improvements, recognise when strategies are working and adjust intervention more quickly. Rather than using a “one-size-fits-all” approach, therapy can become increasingly individualised.

For families, this means therapy may become more responsive to a child’s changing needs.

More opportunities to practise skills

Learning new skills takes repetition, and sometimes there simply isn’t enough time during a weekly therapy session. AI-powered technologies, such as interactive apps or virtual reality (VR), can provide children with additional opportunities to practise skills between sessions.

For example, some VR programs allow children to rehearse everyday situations like starting conversations, ordering food, attending school or participating in a job interview. Because these environments are safe and predictable, children can build confidence before trying these skills in real life.

For many children—particularly autistic children—this structured practice can reduce anxiety and make new experiences feel more manageable.

Better access to services

Families living in rural or remote areas often face significant barriers to accessing specialised services. AI-supported digital tools may help therapists deliver aspects of care remotely, monitor progress between appointments and provide additional support without requiring frequent travel.

While these technologies cannot replace face-to-face therapy, they can improve access for families who might otherwise struggle to receive ongoing care.

What AI cannot do

Despite all the excitement, AI has important limitations. Children are far more than a collection of behaviours or data points. Therapists don’t simply observe what a child does—they seek to understand why they behave that way. A therapist considers relationships, emotions, family experiences, culture, sensory needs, motivation, trauma history and many other factors that AI simply cannot fully appreciate.

A computer might recognise that a child avoids eye contact. A skilled clinician asks whether the child is anxious, overwhelmed, concentrating, culturally respectful, masking, or simply communicating in a different way. That human understanding remains essential.

Are there any risks?

Like any new technology, AI also raises important questions.

Privacy matters

Many AI systems require large amounts of information to function effectively. This may include videos, speech recordings or behavioural information.

Parents should feel confident asking:

  • Who owns my child’s data?
  • How is it stored?
  • Who has access to it?
  • Will it be used to improve future AI systems?

Reputable providers should always explain how information is collected, stored and protected.

AI can make mistakes

AI is only as good as the information it has been trained on. If the data used to develop an AI system doesn’t include children from different cultures, ages, communication styles or support needs, the system may not perform equally well for everyone. This is one reason why AI should always be used alongside professional clinical judgement—not instead of it.

Every child deserves to be themselves

An important discussion in autism research is whether technology encourages children to “look more typical” rather than supporting authentic communication. For example, teaching eye contact may appear socially successful but may also be uncomfortable or exhausting for some autistic people.

Today’s best practice focuses on helping children build meaningful communication, confidence and independence—not simply teaching them to behave like everyone else. The goal should always be to support the child, not to change who they are.

The future of therapy

AI is likely to become a normal part of therapeutic services over the next decade. We may see tools that help therapists monitor progress, identify changes earlier, personalise intervention plans and reduce paperwork, allowing clinicians to spend more time working directly with children and families. This is perhaps where AI offers its greatest benefit.

Rather than replacing therapists, AI can take care of routine administrative and data-processing tasks, giving clinicians more time to build relationships, problem-solve with families and provide compassionate care.

What should parents remember?

AI is an exciting development in healthcare, but it is exactly that—a tool. The most effective therapy still depends on strong relationships, skilled clinical judgement and a deep understanding of each child’s unique strengths and challenges. Technology may help therapists work smarter, but it cannot replace empathy, trust or human connection.

As AI continues to evolve, parents can feel reassured that the best therapeutic care will combine the strengths of both worlds: innovative technology supported by experienced professionals who know that every child is much more than the data they generate.

At the end of the day, AI isn’t here to replace your therapist. It’s here to help your therapist spend more time doing what matters most: supporting your child and your family.

References

This blog is based on a recent assignment completed as part of a Master’s degree in clinical psychology and draws on evidence from the literature it reviews, including:

  • Abbas et al. (2018). Journal of the American Medical Informatics Association.
  • CieÅ›lik et al. (2020). Complementary Therapies in Medicine.
  • Chen et al. (2022). Interactive Learning Environments.
  • Denecke et al. (2021). Artificial Intelligence for Chatbots in Mental Health.
  • Ebert et al. (2019). Frontiers in Psychiatry.
  • Fiske et al. (2019). Journal of Medical Internet Research.
  • Jovanović et al. (2021). Artificial Intelligence: Theory and Applications.
  • Magrabi et al. (2019). Yearbook of Medical Informatics.
  • Wankhede et al. (2024). Asian Journal of Psychiatry.
  • World Health Organization (2024). Clinical Descriptions and Diagnostic Requirements for ICD-11.