Digital twins, often fuelled by Artificial Intelligence (AI) and Cloud, are currently one of the most debated technologies in the public sector. Their potential to disrupt the traditional world of policy-making, empower users with new intelligence and facilitate urban operations have attracted the interest of policymakers, researchers and technologists alike. Although adoption can deliver benefits ranging from faster, more transparent decision-making to more accurate and collaborative operations management, the public sector application of digital twin technology is still in relative infancy. As a result, quantitative and qualitative insights into the implementation and impact of digital twins remain thin, particularly at the local or urban level. This paper examines early use cases alongside implementation approaches and governance tools to identify the benefits and challenges inherent in replacing traditional urban policy-making and planning approaches with Local Digital Twins (LDT) to force multiply or bolster outcomes. It argues that greater multi-stakeholder experimentation, research and collaboration is needed in the urban digital twin space in order to develop effective governance solutions that support and scale more comprehensive public sector transformation aims.