This study investigates how social network analysis (SNA) can enhance risk assessment in Peer to Peer (P2P) lending platforms, where the rapid growth of digital lending highlights the background challenge of accurately evaluating borrower credibility beyond traditional financial metrics. The objective of this research is to develop a network based analytical framework that identifies relational patterns among borrowers and lenders to improve the detection of potential default risks. The method integrates graph based modeling, centrality measurements, and community detection algorithms applied to borrower interaction data, enabling the identification of structural features such as influence, connectivity, and hidden borrower clusters that may correlate with credit behavior. The results demonstrate that borrowers with high betweenness and eigenvector centrality values exhibit significantly different default tendencies compared to those in more isolated network positions, while community structures reveal risk concentrated clusters that are not captured by conventional credit scoring systems. The conclusion emphasizes that incorporating SNA into P2P lending risk assessment provides a more holistic and data driven understanding of borrower behavior, allowing platforms to strengthen credit evaluation, reduce default rates, and support more sustainable digital lending ecosystems.