Building Socially Aware Machines: The New Frontier of AI Theory of Mind

Author: Hajar El Qasemy

Issue: Spring Issue, 2026

Abstract

The research focus was motivated by a curiosity about what lies beyond current artificial intelligence (AI) capabilities and an interest in exploring the next frontier of AI evolution. This research is conducted as a literature review, with the purpose of bringing to light ongoing advances in theory of mind (ToM) AI and outlining future AI directions, specifically for readers who are seeking insights into what’s on the horizon in the field of AI. Concretely, the literature review explores the emergence of ToM in AI, tracing its evolution from traditional AI systems towards ToM models capable of comprehending and predicting human mental states. Through a discussion of the current landscape, challenges, and future directions, this literature review clarifies how close AI is to achieving fully-fledged ToM and what’s beyond the ultimate realization of ToM. This research reviews recent peer-reviewed sources: empirical and theoretical/conceptual journal articles, review papers, book chapters, conference proceedings, preprints, textbooks, and extended abstracts that were published within the last five years. The findings indicated that ToM in AI is still under development, with its current applications being either in research or experimental phases. Excellent progress was observed in areas such as emotion recognition, predictive modeling, conversational AI, multi-agent systems, simulation, and cognitive modeling. However, constructing mental models with ToM capabilities remains challenging, particularly in leveraging meta-learning to accurately represent existing intelligent entities, whether artificial or human. The conclusion showed that our world still does not fully grasp complex human thoughts, and creating AI systems that can adapt to our ever-changing minds and infer internal mental states is an ambitious milestone beyond which lies self-awareness, a drastic shift in technology that the world may not be ready for.