Artificial intelligence (AI) is increasingly transforming logistics operations, offering potential improvements in efficiency, decision-making, and responsiveness. However, despite growing investments, many organisations struggle to move from experimentation to large-scale implementation. This study aims to explore how organisations adopt and scale AI in logistics by analysing the process through the lens of dynamic capabilities, focusing on sensing, seizing,and transforming. The study is based on a qualitative research design, using 11 semi-structured interviews with senior professionals across logistics-related organisations in Sweden, Germany, and the United States. The findings are further supported by secondary data sources to strengthen the analysis. The findings show that AI adoption is uneven across stages. While organisations are strong in sensing opportunities through experimentation, they face challenges in seizing and especially transforming AI initiatives into integrated and scalable solutions. Key barriersinclude organisational silos, lack of trust, data governance issues, and limited cross-functional coordination. The study highlights that successful AI adoption depends on organisational capabilities rather than technology alone.