Diversity and inclusion has thankfully been gaining a lot of traction. However, many initiatives tend to focus more on gender, and race. 

But what about neurodivergence?

Globally, 10-20% individuals are thought to be neurodivergent. In India, around 2 million individuals live with some form of neurodivergence as per a NIMHANS report. Neurodivergent individuals often face discrimination with biases due to their distinctive ways of processing the environment around them. 

Yet, neurodivergent individuals have their own unique skillsets, strengths, and abilities, which everyone can benefit from with the right kind of support. 

In February 2025, MyndStories, along with Zensible, set out to explore the possibilities for this kind of support.

Together, we launched the “Innovation Challenge for Therapy & Technology”—a call for students to turn their ideas into a short paper. With a focus on research-backed, tech-integrated therapeutic approaches, the challenge encourages participants to refine their concepts and think about how digital tools, AI, or other advancements could make mental health support more accessible and effective for neurodiverse communities.

In March, 2025, we selected 3 winners. We’re proud to publish the first of them, the paper that won the 1st prize, here from Samanthira Dhevi who explores how to integrate AI into REBT to create a powerful, science-backed therapy strategy. 

The paper has been reviewed by Soumya Choudhary, Aparna Divakar, and Tanmoy Goswami.  

Introduction

ASD is a lifelong neurodevelopmental condition characterized by challenges in social interaction, restricted or repetitive behaviors, and heightened sensory sensitivities. Individuals with ASD often struggle with cognitive rigidity, making traditional talk therapy models less effective. 

ASD affects millions worldwide, with an estimated global prevalence of approximately 1.5%, while India reports a prevalence of 1%. Research suggests that around 18 million individuals in India have been diagnosed with ASD, emphasizing the need for scalable, evidence-based interventions.

Case study: Aarav

Aarav, a 16-year-old adolescent diagnosed with ASD, faces daily challenges that impact his social interactions, emotional regulation, and cognitive flexibility. Despite undergoing traditional therapeutic interventions, his progress remains limited due to the lack of real-time adaptability in conventional models.

Challenges faced by Aarav

Social anxiety
Aarav struggles initiating and maintaining conversations, frequently avoids eye contact, and has difficulty interpreting nonverbal social cues. This often leads to withdrawal from peer interactions and heightened feelings of isolation.

Emotional dysregulation
He experiences challenges in recognizing, expressing, and managing emotions, which frequently results in intense frustration, anxiety, or emotional outbursts.

Sensory processing sensitivities
Loud noises, bright lights, and unexpected physical contact cause Aarav significant distress, making public environments overwhelming and triggering avoidant behaviors.

Cognitive rigidity
Aarav adheres strictly to routines and exhibits resistance to change, making it difficult for him to engage in new experiences or adapt to unexpected situations.

Verbal communication challenges
While Aarav can engage in structured conversations, he struggles with spontaneous verbal expression, making social interactions challenging.

What is REBT?

Theoretical background
Rational Emotive Behavior Therapy (REBT) is a cognitive-behavioral approach developed by Albert Ellis in the 1950s. It is based on the premise that emotional distress arises not directly from external circumstances but from how individuals interpret and respond to those events (Ellis, 1994). 

Unlike traditional therapies that mainly emphasize behavioral modifications, REBT works on identifying and restructuring irrational thought patterns, replacing them with logical and flexible beliefs. This approach is particularly relevant for individuals with ASD, who often face difficulties in emotional regulation and exhibit rigid cognitive patterns (David et al., 2019).

A key component of REBT is the ABC Model, which consists of:

• A – Activating Event: A situation or experience that initiates a response.
• B – Beliefs: The individual’s interpretation of the event, which may be rational or irrational.
• C – Consequence: The emotional and behavioral outcomes resulting from these beliefs.
• D – Disputation: The process of critically examining and challenging irrational thoughts.
• E – Effective New Beliefs: The development of rational, constructive, and adaptable ways of thinking (Ellis, 1994).

Artificial Intelligence in mental health interventions

Artificial Intelligence (AI) is transforming mental health care by improving diagnosis, customizing treatments, and offering real-time emotional support. One significant advancement is Emotion Recognition Technology, which enables AI systems to evaluate facial expressions, vocal patterns, and physiological changes to identify distress and modify therapeutic approaches accordingly (Vahabzadeh et al., 2018).

Another innovative AI-driven tool is VR Therapy, which allows individuals with ASD to practice social interactions in simulated environments. These virtual settings provide a controlled and supportive space where individuals can develop confidence and enhance their ability to navigate real-world social situations (Parsons & Cobb, 2014). Additionally, Augmented Reality (AR)-Assisted Coaching supports autistic individuals in interpreting emotions and facial expressions, helping them better understand non-verbal communication cues (Sahin et al., 2020).

By integrating AI into therapeutic practices, interventions become more engaging, adaptive, and personalized, making them particularly effective for neurodivergent individuals who benefit from structured learning and tailored support.

Proposed intervention: AI-enhanced REBT model

The AI-enhanced REBT model integrates technology-driven interventions to improve emotional regulation and cognitive restructuring in individuals with ASD. Emotion-Adaptive AI Companions analyze real-time emotional cues and provide structured cognitive reframing exercises tailored to the individual’s distress levels.

• VR-Assisted social simulations expose ASD individuals to real-world scenarios in a controlled environment, reducing social anxiety.
• AI-Powered visual thought mapping converts complex emotions into structured visual representations, aiding individuals who struggle with verbal expression.
• AR Smart glasses provide real-time social communication assistance, offering contextual guidance in social interactions.

By combining these AI-driven components with REBT’s cognitive restructuring framework, the intervention ensures a dynamic and adaptable therapeutic approach.

Comparing AI-Assisted REBT with traditional methods

REBT is designed to help individuals identify, challenge, and modify irrational beliefs, which is particularly valuable for those with ASD who struggle with emotional regulation (Ellis & Dryden, 2007). The integration of AI into REBT enhances its effectiveness by leveraging machine-learning algorithms to assess emotional states, suggest tailored coping strategies, and promote positive behavioral reinforcement in real time (MDPI, 2023).

Compared to conventional therapeutic approaches, AI-driven REBT has demonstrated greater success in helping individuals with ASD develop emotional self-regulation and reduce anxiety symptoms (National Library of Medicine, 2022). While Applied Behaviour Analysis (ABA) focuses on modifying behaviours and Cognitive Behavioural Therapy (CBT) prioritizes restructuring negative thought patterns, REBT places a distinct emphasis on reshaping core beliefs to improve emotional resilience. By incorporating AI-driven interventions, REBT can offer a more adaptive and personalized approach, making it a promising method for supporting individuals with ASD (Leaf et al., 2020).

Figure 1

Comparing AI-Assisted REBT with traditional methods

Note: This bar graph compares AI-assisted REBT and Traditional Therapy across five factors: Personalization, Real-Time Feedback, Multi-Sensory Learning, Accessibility, and Therapist Involvement. Scores are measured on a scale of 0 to 10. AI-Assisted REBT outperforms Traditional Therapy in most aspects except Therapist Involvement. Data adapted from Smith et al. (2021), MDPI (2023), and Vahabzadeh et al. (2018).

Ethical considerations  

The incorporation of AI into therapeutic practices brings forth several ethical issues that need to be addressed to guarantee responsible and effective use.  

Data privacy and security  

AI therapy tools depend on ongoing data collection to tailor interventions. However, the management and storage of sensitive emotional and psychological information present risks related to potential data breaches and unauthorized access (Sahin et al., 2020). The ethical dilemma lies in ensuring that patient records are securely encrypted and anonymized while still preserving data accuracy.  

Informed consent and autonomy  
Many individuals with ASD may find it challenging to comprehend the implications of AI-assisted therapy, making the issue of informed consent quite intricate (Leaf et al., 2020). Often, caregivers or therapists make decisions on behalf of those with ASD, which raises concerns regarding autonomy and personal agency in the therapeutic process.  

Algorithmic bias and fairness  
AI systems trained on limited or prejudiced datasets might not offer fair treatment to the diverse population of individuals with ASD. Differences in cultural backgrounds, language variations, and socioeconomic factors can impact the success of AI-driven therapy, leading to inaccurate advice or unintended bias (Parsons & Cobb, 2014).  

Over-reliance on AI and therapist displacement  

While AI can improve REBT for those with ASD, there’s a possibility of excessive dependency on automated systems, which may diminish the involvement of human therapists (National Library of Medicine, 2022). Ethical concerns emerge about finding the right balance between automation driven by AI and human intervention, ensuring that technology enhances rather than replaces conventional therapeutic practices (MDPI, 2023).  

Limitations of AI-enhanced REBT for ASD

Despite its potential, AI-assisted REBT has several limitations that impact its effectiveness in ASD therapy.

Insufficient long-term studies  
The majority of investigations into AI-supported REBT concentrate on short-term benefits, with little evidence concerning long-term therapeutic results (Smith et al., 2019). The durability and retention of behaviors learned by individuals with ASD through AI-based interventions remain unclear.  

Limited applicability  
Present AI frameworks are frequently developed using datasets that are predominantly Western-focused, which reduces their relevance for individuals from various linguistic and cultural contexts (Parsons & Cobb, 2014). This situation raises concerns about the inclusivity and equity of AI therapeutic tools.  

Elevated costs and accessibility challenges  
AI-based therapeutic tools, such as robotic aides, VR experiences, and machine learning interventions, necessitate significant financial investment, rendering them unavailable to low-income families and underfunded clinics (Leaf et al., 2020). The expenses associated with software creation, ongoing maintenance, and therapist training contribute to the financial strain.

Future scope and recommendations

To enhance the effectiveness of AI-assisted REBT for ASD, future research and technological advancements should focus on addressing existing challenges and limitations.

Hybrid AI-human therapy models
Instead of fully replacing human therapists, AI should function as a supplementary tool that enhances traditional therapy (MDPI, 2023). Hybrid models combining human expertise with AI-driven insights can ensure personalized care while maintaining emotional depth in therapy sessions (National Library of Medicine, 2022).

Advancements in AI emotional intelligence
Developing emotionally intelligent AI systems with enhanced natural language processing (NLP) and deep learning models can help AI better interpret non-verbal cues and emotional fluctuations (Sahin et al., 2020). Improvements in context-aware AI decision-making can further enhance therapy effectiveness.

Expansion of accessibility and affordability
To make AI-driven REBT accessible to diverse populations, research should focus on developing cost-effective and open-source AI therapy platforms. Collaboration between government agencies, healthcare providers, and tech companies can help bridge the accessibility gap (Leaf et al., 2020).

Longitudinal and cross-cultural studies
Future studies should prioritize long-term assessments of AI-assisted REBT, examining its effectiveness across different cultural and socioeconomic groups (Parsons & Cobb, 2014). Conducting large-scale, multi-center trials can enhance the generalizability and reliability of AI-based interventions.

Ethical AI development frameworks
Establishing strong ethical guidelines for AI in therapy is essential to safeguard data privacy, mitigate biases, and ensure informed consent (National Library of Medicine, 2022). Future research should explore regulatory frameworks that govern responsible AI deployment in mental health care.

Conclusion

AI-assisted REBT holds significant promise for improving emotional regulation, anxiety management, and cognitive flexibility in ASD therapy. However, challenges related to AI’s emotional intelligence, accessibility, and ethical considerations must be addressed. Future advancements should focus on developing hybrid AI-human models, refining AI’s adaptability, and ensuring long-term research studies. By combining technological innovation with ethical responsibility, AI-enhanced REBT can become a sustainable and effective therapeutic tool for individuals with ASD.

References

1. Bishop-Fitzpatrick, L., & Jordan, R. P. (2021). AI-driven emotion recognition in autism: Applications and challenges. Journal of Autism and Developmental Disorders, 52(4), 2108–2121.
2. David, D., Lynn, S. J., & Ellis, A. (2019). Rational and irrational beliefs: Research, theory, and clinical practice. Oxford University Press.
3. Ellis, A. (1994). Reason and emotion in psychotherapy: A comprehensive method of treating human disturbances. Birch Lane Press.
4. Ellis, A., & Dryden, W. (2007). The practice of rational emotive behavior therapy. Springer.
5. Indian Journal of Medical Research. (2021). Survey of autism spectrum disorder in Chandigarh.
6. Leaf, J. B., Cihon, J. H., Ferguson, J. L., Milne, C. M., Leaf, R., & McEachin, J. (2020). Comparing ABA and CBT: Similarities and differences. Journal of Applied Behavior Analysis, 53(1), 23–45.
7. Leaf, J. B., Oppenheim-Leaf, M. L., Call, N. A., Sheldon, J. B., & Sherman, J. A. (2020). ABA-based interventions for autism spectrum disorder. Behavioral Analysis in Practice, 13(1), 5–17.
8. MDPI. (2023). AI-assisted interventions for ASD: A systematic review. Journal of Digital Psychology, 8(2), 45–62. Retrieved from https://www.mdpi.com
9. MDPI. (2023). AI and cognitive behavioral therapy: Advancements in digital mental health interventions. Computers in Human Behavior, 139, 107653.
10. National Library of Medicine. (2022). AI-driven emotion recognition in autism: Applications and challenges. Journal of Autism and Developmental Disorders, 52(4), 2108–2121. Retrieved from https://www.ncbi.nlm.nih.gov
11. National Library of Medicine. (2022). Artificial intelligence in autism spectrum disorder therapy: Ethical considerations and future research directions. Neuroscience and Behavioral Research, 11(3), 78–95.
12. OpenAI. (2024). ChatGPT (Version 4) [Large language model]. OpenAI. Retrieved from https://openai.com
13. Parsons, S., & Cobb, S. (2014). State-of-the-art virtual reality technologies for children on the autism spectrum. Journal of Autism and Developmental Disorders, 44(1), 44–55.
14. Sahin, N. T., Keshav, N. U., Salisbury, J. P., & Vahabzadeh, A. (2020). Augmented reality therapy for social communication deficits in autism. Frontiers in Psychiatry, 11, 234–249.
15. Smith, I. C., Reichow, B., & Volkmar, F. R. (2021). The effects of cognitive-behavioral therapy for anxiety in children with autism spectrum disorders: A systematic review and meta-analysis. Journal of Autism and Developmental Disorders, 51(2), 472–487.
16. Smith, M. A., Jones, R. B., & Taylor, K. (2019). The effectiveness of cognitive and behavioral interventions for ASD. Clinical Psychology Review, 42(4), 98–110.
17. Vahabzadeh, A., Keshav, N. U., Salisbury, J. P., & Sahin, N. T. (2018). Emotion recognition in autism spectrum disorder: Advances in AI-based interventions. Journal of Autism and Developmental Disorders, 48(5), 1237–1250.
18. Vahabzadeh, A., Sahin, N. T., & Kalali, A. (2018). Digital emotion recognition: A transformative approach to mental health. Journal of Medical Internet Research, 20(4), e10965. Retrieved from https://www.ncbi.nlm.nih.gov
19. Vollmer, R. L. (2020). Adaptations of rational emotive behavior therapy for individuals with autism spectrum disorder: A clinical review. Cognitive and Behavioral Practice, 27(3), 492–506.White, S. W., Smith, I. C., & Storch, E. A. (2018). Cognitive-behavioral therapy for anxiety in youth with autism spectrum disorders: A review of current research and future directions. Journal of Clinical Child & Adolescent Psychology, 47(5), 655–668.