Towards Explainable Public Sector AI: An Exploration of Neuro-Symbolic AI and Enterprise Modeling (Short Paper)

Artificial Intelligence (AI) offers transformative potential for enhancing public sector services. However, the lack of explainability within many AI systems, particularly those relying on opaque ‘black box’ neural networks, hinders widespread adoption due to concerns about fairness, accountability, and trust. To address this, explainable AI (XAI) has emerged as a vital area of research, aiming to illuminate the reasoning behind AI decisions. Neuro-Symbolic AI (NSAI) provides a powerful technique for XAI, combining the strengths of symbolic reasoning with neural pattern recognition. Enterprise Modeling (EM) offers systematic methodologies to capture multifaceted domain knowledge, including process models, rules, and ontologies. This paper explores the potential synergy between XAI, NSAI, and EM for developing trustworthy and transparent AI solutions tailored to the public sector. While this integration offers exciting possibilities for explainable AI, the field is still nascent.