Building Emotional Intelligence into Digital Therapy AI Agents through Neurofeedback

Sam Nallaperuma-Herzberg, Rishabh Balse, Sonia Koszut, Lilith Stenhouse, Anna Bevan, Tristan Bekinschtein, Pietro Lio
Proceedings of the First Workshop on NeuroAI Multimodal Intelligence @ AAAI 2026, PMLR 308:149-154, 2026.

Abstract

We present a novel emotionally intelligent agent framework for delivering cognitive behavioural therapy (CBT). The system aggregates text sentiment cues with neurofeedback, yielding a fine-grained perception of user state building empathy into the agent. A reinforcement learning (RL) planner maps this affective state to appropriate therapeutic acts, which are verbalised by a large language model (LLM). To enhance reliability, the LLM agent is augmented with a meta-cognitive control layer that continuously self-monitors and refines of its responses. In preliminary studies, the proposed system has demonstrated improved therapeutic efficacy over standard LLM-based agents, as measured by standard psychotherapy metrics. These results highlight the potential of combining neurofeedback, affective computing, RL decision making, and LLM generation to deliver clinically meaningful, scalable CBT paving the way for safe, personalised mental health support at population scale.

Cite this Paper

BibTeX
@InProceedings{pmlr-v308-nallaperuma-herzberg26a, title = {Building Emotional Intelligence into Digital Therapy AI Agents through Neurofeedback}, author = {Nallaperuma-Herzberg, Sam and Balse, Rishabh and Koszut, Sonia and Stenhouse, Lilith and Bevan, Anna and Bekinschtein, Tristan and Lio, Pietro}, booktitle = {Proceedings of the First Workshop on NeuroAI Multimodal Intelligence @ AAAI 2026}, pages = {149--154}, year = {2026}, editor = {Abbasi-Asl, Reza and Iqbal, Asim and Ito, Shinya and Arkhipov, Anton and Sanborn, Sophia}, volume = {308}, series = {Proceedings of Machine Learning Research}, month = {27 Jan}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v308/main/assets/nallaperuma-herzberg26a/nallaperuma-herzberg26a.pdf}, url = {https://proceedings.mlr.press/v308/nallaperuma-herzberg26a.html}, abstract = {We present a novel emotionally intelligent agent framework for delivering cognitive behavioural therapy (CBT). The system aggregates text sentiment cues with neurofeedback, yielding a fine-grained perception of user state building empathy into the agent. A reinforcement learning (RL) planner maps this affective state to appropriate therapeutic acts, which are verbalised by a large language model (LLM). To enhance reliability, the LLM agent is augmented with a meta-cognitive control layer that continuously self-monitors and refines of its responses. In preliminary studies, the proposed system has demonstrated improved therapeutic efficacy over standard LLM-based agents, as measured by standard psychotherapy metrics. These results highlight the potential of combining neurofeedback, affective computing, RL decision making, and LLM generation to deliver clinically meaningful, scalable CBT paving the way for safe, personalised mental health support at population scale.} }
Endnote
%0 Conference Paper %T Building Emotional Intelligence into Digital Therapy AI Agents through Neurofeedback %A Sam Nallaperuma-Herzberg %A Rishabh Balse %A Sonia Koszut %A Lilith Stenhouse %A Anna Bevan %A Tristan Bekinschtein %A Pietro Lio %B Proceedings of the First Workshop on NeuroAI Multimodal Intelligence @ AAAI 2026 %C Proceedings of Machine Learning Research %D 2026 %E Reza Abbasi-Asl %E Asim Iqbal %E Shinya Ito %E Anton Arkhipov %E Sophia Sanborn %F pmlr-v308-nallaperuma-herzberg26a %I PMLR %P 149--154 %U https://proceedings.mlr.press/v308/nallaperuma-herzberg26a.html %V 308 %X We present a novel emotionally intelligent agent framework for delivering cognitive behavioural therapy (CBT). The system aggregates text sentiment cues with neurofeedback, yielding a fine-grained perception of user state building empathy into the agent. A reinforcement learning (RL) planner maps this affective state to appropriate therapeutic acts, which are verbalised by a large language model (LLM). To enhance reliability, the LLM agent is augmented with a meta-cognitive control layer that continuously self-monitors and refines of its responses. In preliminary studies, the proposed system has demonstrated improved therapeutic efficacy over standard LLM-based agents, as measured by standard psychotherapy metrics. These results highlight the potential of combining neurofeedback, affective computing, RL decision making, and LLM generation to deliver clinically meaningful, scalable CBT paving the way for safe, personalised mental health support at population scale.
APA
Nallaperuma-Herzberg, S., Balse, R., Koszut, S., Stenhouse, L., Bevan, A., Bekinschtein, T. & Lio, P.. (2026). Building Emotional Intelligence into Digital Therapy AI Agents through Neurofeedback. Proceedings of the First Workshop on NeuroAI Multimodal Intelligence @ AAAI 2026, in Proceedings of Machine Learning Research 308:149-154 Available from https://proceedings.mlr.press/v308/nallaperuma-herzberg26a.html.

Related Material