Self-Explainability in Self-Adaptive and Self-Organising Systems: Status and Research Directions

A systematic literature review examines Self-Explainability (SX) in self-adaptive and self-organizing systems, finding that most approaches remain theoretical with no standardized evaluation methods. The research establishes a taxonomy and framework for advancing SX, identifying a significant gap between conceptual work and practical implementation in increasingly complex AI-driven systems.

This research addresses a critical challenge in AI system governance: as autonomous systems become more sophisticated, understanding their decision-making processes becomes essential for trust and accountability. Self-Explainability represents an evolution beyond traditional Explainable AI, pushing systems to autonomously clarify their behavior rather than requiring external interpretation. The distinction matters significantly for real-world deployment, where systems must operate with minimal human oversight.

The findings reflect broader industry tensions between capability and comprehension. Complex self-organizing systems drive innovation across sectors—autonomous vehicles, financial trading algorithms, and network management—yet their opacity creates regulatory and safety concerns. Organizations increasingly face pressure from stakeholders to justify algorithmic decisions, making self-explanation a competitive advantage rather than optional feature.

For developers and enterprises, this review consolidates fragmented research into actionable frameworks, identifying that current implementations lack rigor in evaluation. The absence of formal standards creates implementation ambiguity and inconsistent quality across projects. This gap represents both risk and opportunity: organizations investing in robust SX frameworks gain credibility and compliance advantages, while those relying on immature approaches face potential regulatory scrutiny.

The research roadmap suggests SX development will accelerate as regulatory bodies demand greater AI transparency. Organizations building self-adaptive systems should prioritize explainability architecture early rather than retrofitting solutions. The establishment of evaluation standards—likely within 2-3 years—will reshape industry practices and potentially create new compliance requirements for deployed systems.

• →Most Self-Explainability approaches remain conceptual with few production implementations, indicating early-stage market development.
• →No standardized evaluation methods exist for assessing Self-Explainability, creating implementation uncertainty for developers.
• →Self-Explainability enables systems to autonomously justify decisions, advancing beyond traditional Explainable AI frameworks.
• →Regulatory pressure for AI transparency will likely accelerate SX adoption as a compliance requirement.
• →Early investment in SX architectures provides competitive advantages as industry standards emerge.