Accelerating permafrost thaw and coastal erosion in rural Alaska destabilize the built environment, increasing the risk of sociotechnical failures that negatively impact nearby communities. Infrastructure adaptation is necessary to mitigate these threats, yet epistemic uncertainty remains about where the most exposed infrastructure is located and the corresponding community impact of failure. As a result, existing adaptation activity may not be prioritized according to relative need, potentially reducing the efficiency and effectiveness of adaptation activity. To address this gap, this study seeks to identify the likely failures, cascading impacts, as well as which communities are likely to experience them. To do so, this study employs machine learning techniques and permafrost terrain maps to identify vulnerable coastlines. Using density-based cluster mapping, statistical summarization, and semi-cognitive mapping, we explore the locations and functions of infrastructure, and infrastructure interdependencies occurring on Alaskan coastlines. Results suggest the following to improve vulnerable infrastructure outcomes: (1) allocate public funding to support adaptation of exposed infrastructure, (2) increase financial and physical resources for resilience research and development for exposed infrastructure, and (3) expand access to adaptation activity through increased local input in decision-making. For instance, as results show that water and sewer infrastructure in the Bethel census area is highly exposed, results could support additional funding to organizations that repair and relocate such infrastructure. Through these changes, regional adaptation activity can become more prioritized – ensuring that funds address the most pressing needs.