The rapid adoption of artificial intelligence in the last decade has transformed numerous sectors, including healthcare, manufacturing, finance, and logistics. As a foundational discipline supporting these industries, project management has a significant potential to benefit from this technological evolution. In particular, deep learning, a subset of artificial intelligence, is transforming project management by equipping its practitioners with advanced analytical tools that enable data-driven decision-making, leading to improved efficiency, adaptability, and success. This study explores the state of art of deep learning applications in project management, with a focus on the benefits and challenges, to identify knowledge gaps in utilizing these techniques to enhance project outcomes. Our findings highlight two primary knowledge gaps: first, the lack of empirical research exploring how deep learning integration is reshaping project managers' responsibilities, despite its growing adoption; and second, the limited application of artificial intelligence across all project phases, especially initiation and closure. While deep learning has the potential to improve overall project performance, challenges such as data integrity, ethical considerations, and model transparency remain unaddressed. Overcoming these obstacles requires closer collaboration between researchers and project managers to develop practical, specific solutions. In this context, this study emphasizes the need for further research to explore the full impact of deep learning techniques on project management and to bridge the identified gaps.