Autonomous Circular Economy Systems: The Role of AI Agents and Digital Twins in Self-Optimizing Sustainable Business Ecosystems

The study establishes that AI Agent Autonomy and Digital Twin Capability drive Sustainable Business Performance indirectly through a sequential resource–capability–process–performance pathway.

• Mechanism: These technologies enhance Self-Optimization Capability, which enables Circular Process Integration, while Algorithmic Trust serves as a critical behavioral governance factor.
• Key Finding: Direct links between the technologies and performance are weak or non-significant; value is created only when digital resources are converted into organizational routines and stakeholder trust.
• Framework: Grounded in Resource-Based View and Dynamic Capabilities Theory, the model explains how sensing, seizing, and reconfiguring resources via AI and digital twins embeds circular principles into operations.

The article examines how autonomous digital technologies—particularly AI agents and digital twins—can be combined into circular-economy systems that continuously optimize resource flows, asset lifecycles, and service delivery. It grounds this argument in the Resource-Based View and dynamic-capabilities framework, treating AI-enabled sensing, prediction, decision-making, and simulation as strategic resources and capabilities that allow firms to identify circular opportunities, reconfigure processes, and adapt to changing market, regulatory, or environmental conditions. In this framing, digital twins serve as virtual representations of physical assets, supply chains, and material loops, while AI agents coordinate decisions across operational layers, enabling closed-loop recovery, remanufacturing, recycling, and more efficient product-service systems.

A key insight is that circularity is not merely a design or operational target but can become an emergent property of self-optimizing digital ecosystems. AI agents can monitor asset condition, forecast degradation or residual value, recommend circular interventions, and coordinate actions across supply-chain actors; digital twins can simulate alternative configurations, quantify trade-offs, and support continuous learning from operational and environmental feedback. The article’s contribution is therefore to provide a conceptual architecture that connects autonomous digital infrastructure to sustainable business performance, emphasizing that value creation depends on integrated data, adaptive control, and ecosystem-level governance rather than isolated digital tools.

The material matters because it offers a theoretically grounded rationale for designing and scaling AI-driven circular operations. For a technically literate audience, the value lies in shifting from static circularity metrics to adaptive, data-intensive control loops that balance environmental, economic, and social objectives in real time. It also implicitly highlights the enabling conditions and risks involved: data quality, interoperability, cybersecurity, trust, and governance structures that determine whether autonomous systems reinforce sustainable outcomes or merely optimize narrow performance targets.