The increasing electrification of transport and heating sectors through electric vehicles and heat pumps presents significant

challenges for low-voltage distribution grids, leading to potential congestion situations. We present a decentralized agent-based

optimization approach for congestion management that leverages both legally mandated consumer flexibility and voluntary battery

flexibility. This multi-agent system enables individual households to participate in grid congestion management through

autonomous decision-making while aiming to preserve household data sovereignty and to achieve privacy-preserving coordination

via minimal information exchange. Our approach incorporates the German Energy Industry Act (§ 14a EnWG) framework

and utilises the Alternating Direction Method of Multipliers (ADMM) for decentralized optimization. Evaluation results using

the IEEE-33 test system demonstrate significant peak load reduction, complete elimination of thermal constraint violations, and

real-time feasibility. These findings demonstrate that decentralized coordination can effectively manage grid congestion while

maintaining privacy and regulatory compliance, offering a scalable alternative to traditional centralized approaches for future

smart grid implementations.