<p>The logistics industry is crucial to global trade, but it is under increasing pressure to reduce its environmental impact and improve sustainability. Artificial Intelligence (AI) and Machine Learning (ML) have become promising tools for optimizing logistics operations, but their incorporation into sustainability frameworks remains insufficiently explored. This study conducts a systematic review of 71 peer-reviewed articles to examine how AI/ML techniques are applied in sustainable logistics, with a focus on their alignment with the economic, environmental, and social aspects of the triple bottom line. Using Porter’s Value Chain and the Sustainability Balanced Scorecard as analytical lenses, we categorize applications across logistics activities and differentiate between direct and indirect sustainability impacts. Our findings reveal a fragmented understanding of sustainability, a methodological imbalance that favors machine learning over hybrid approaches, and a heavy reliance on synthetic datasets with a strong geographic bias toward Asia. Although promising results are evident, especially when policy mechanisms such as carbon caps or taxes are incorporated, the field remains nascent. We conclude that achieving comprehensive sustainability in logistics requires innovative methods, the incorporation of real-world data, and closer alignment with regulatory and social priorities.</p>