1 Introduction
In recent years, modern supply chains have been increasingly characterized by uncertainty arising from fluctuating demand, geopolitical instability, and disruptions caused by events such as pandemics and natural disasters (
Ning et al., 2023;
Li et al., 2024b). These uncertainties, coupled with the growing complexity of supply chain networks, often lead to inefficiencies, higher costs, and delays in decision-making (
Xue and Li, 2023;
Bednarski et al., 2025). Efficiency challenges, such as inventory mismanagement, transportation delays, and suboptimal resource allocation, remain prevalent (
Xue and Li, 2023;
Li et al., 2024a). Moreover, data silos between supply chain partners hinder collaboration, reducing transparency and operational coherence (
Bednarski et al., 2025). Traditional supply chain management (SCM) struggles to address forecasting uncertainties due to dynamic market conditions and the lack of real-time data integration. These issues have driven the evolution of supply chain strategies, making advanced technologies essential for managing and optimizing supply chain processes (
Sharma et al., 2022; Wu et al., 2024;
Dolgui and Ivanov, 2025).
Among the transformative technologies shaping modern supply chains, Artificial Intelligence (AI) has gained significant attention (
Li and Li, 2022;
Hendriksen, 2023;
Chen et al., 2024). Within the broader AI landscape, Generative AI (GAI) has recently emerged as a groundbreaking innovation with the potential to revolutionize SCM (
Wamba et al., 2023;
Jackson et al., 2024). Powered by advanced algorithms such as transformer models, GAI can synthesize new data, generate human-like responses, and model complex scenarios (
Floridi and Chiriatti, 2020;
Dubey et al., 2024;
Li et al., 2024b). Unlike traditional AI, which primarily focuses on analyzing historical data or performing specific tasks, GAI excels in creating novel solutions, simulating potential outcomes, and offering creative approaches to problem-solving (
Budhwar et al., 2023;
Fosso Wamba et al., 2024). These capabilities position GAI as a game-changer for addressing the multifaceted challenges faced by supply chain managers (
Li et al., 2024a). According to statistical analysis, the global GAI in supply chain market is projected to reach USD 10,284 million by 2032, up from USD 269 million in 2022, growing at a compound annual growth rate of 45.3% during the forecast period from 2023 to 2032, as shown in Fig.1.
To date, despite the promising potential of GAI, its application within SCM remains in its early stages. Existing research in AI for SCM has primarily concentrated on predictive and prescriptive analytics, while the creative and generative capabilities of AI remain largely unexplored. Most studies have focused on incremental improvements rather than the transformative changes that GAI could introduce (
Wamba et al., 2023;
Li et al., 2024a and
2024b). For example, although traditional AI models have effectively optimized inventory levels and forecasted demand fluctuations, they often lack the capacity to generate innovative strategies for mitigating disruptions or redesigning supply chain networks. This gap underscores the limited theoretical integration of GAI within SCM. The existing literature lacks a unified theoretical framework that systematically explores the integration of GAI into key supply chain functions such as demand forecasting, inventory optimization, and risk management. Furthermore, the literature has not sufficiently examined how GAI can be leveraged to optimize decision-making processes in real-time, considering the dynamic nature of supply chains. To address this gap, the research will focus on the following key questions: (i) What are the core GAI capabilities necessary for building the SCM framework? (ii) What are the potential benefits and limitations of integrating GAI into supply chain processes? (iii) What knowledge gaps exist in the current literature?
To examine the above questions, our study employs a conceptual integration method. By synthesizing relevant theories of GAI and SCM with practical cases, a comprehensive theoretical framework is developed. Specifically, we first introduce the background and significance of GAI, and focus on the core capabilities and outlines their applications in SCM. Then, we illustrate the empowerment mechanisms and challenges of GAI in SCM and propose corresponding solutions. Afterward, we establish a research agenda that identifies key questions and avenues for future exploration. By addressing these aspects, this study seeks to contribute to the growing body of knowledge on Generative AI and its transformative role in SCM, while providing a roadmap for its effective implementation in practice.
2 Core capabilities of GAI
The increasing complexity of global supply chains has presented critical challenges for businesses in coordinating resources, forecasting demand, and dynamically optimizing processes. GAI’s unique capabilities provide innovative solutions to these challenges. This section seeks to investigate the core GAI capabilities necessary for building a comprehensive SCM framework. The analysis presented in this study builds upon the capabilities described in previous literature, extending their proposals (
Floridi and Chiriatti, 2020;
Budhwar et al., 2023;
Wamba et al., 2023;
Dubey et al., 2024;
Feuerriegel et al., 2024;
Fosso Wamba et al., 2024;
Jackson et al., 2024;
Li et al., 2024a and
2024b;
Modgil et al., 2025). The core capabilities of GAI can be classified into five categories: learning and creativity, perception and prediction, expression and communication, collaboration and interaction, and adjustment and adaptation. Tab.1 defines these core capabilities and outlines their applications within SCM.
2.1 Learning and creativity
GAI can extract underlying patterns and features from large-scale, multimodal data, including text, images, and audio. Through semi-supervised or unsupervised learning, GAI can uncover the intrinsic structure and distribution of data, even in the absence of explicit labels, thereby facilitating more efficient decision-making processes. Furthermore, GAI exhibits considerable creative potential. By learning patterns from training data, GAI can generate novel and innovative content. This creative capacity holds significant promise for applications in SCM. Specifically, GAI can predict demand based on real-time data and dynamically optimize solutions for inventory management and logistics, thereby enhancing both the responsiveness and efficiency of the supply chain.
2.2 Perception and prediction
In comparison to traditional AI, GAI offers superior processing and analytical capabilities for images and videos, enabling the extraction of valuable insights from these data sources, thereby enhancing prediction accuracy. Furthermore, advancements in natural language processing, particularly through models such as GPT, have significantly broadened AI’s perceptual capabilities. Once trained, GAI models can understand a broad range of text with near-human accuracy, both in terms of semantic and syntactic comprehension. In the context of SCM, these enhanced perceptual and predictive capabilities support more accurate demand forecasting, fostering more responsive and data-driven decision-making.
2.3 Expression and communication
GAI can mimic human thought processes and logic to engage in dialogs, and it also has access to extensive professional knowledge, enabling real-time communication with users. Moreover, GAI supports multiple output formats, including text, images, audio, and video, making it a versatile tool for facilitating expression and communication in diverse contexts. Leveraging these capabilities, GAI enhances the efficiency of interactions between individuals and AI. In the context of SCM, GAI’s expressive and communicative abilities can strengthen communication and collaboration between organizations. Specifically, in supplier relationship management, GAI can assist companies in negotiations, contract formulation, and even optimize contract terms while improving communication efficiency, thereby enhancing both coordination and flexibility within the supply chain.
2.4 Collaboration and interaction
GAI models, such as ChatGPT, demonstrate advanced collaborative and interactive capabilities. These models facilitate real-time dialogs with users, enabling human-machine interaction, and efficiently process large volumes of repetitive tasks, thereby saving valuable team time. For instance, GAI can function as a customer service agent, responding to customer inquiries. GAI can also collaborate with robots in environments like manufacturing, where it handles design and planning, while robots execute specific tasks, thus completing the production process. Additionally, multiple GAI models can operate in tandem, coordinating seamlessly across systems. By leveraging the collaborative and interactive capabilities of GAI, these models can optimize coordination across various stages of the supply chain, thereby enhancing overall efficiency in complex environments.
2.5 Adjustment and adaptation
GAI’s adaptability is primarily evident in its capacity to analyze large data sets, swiftly detect changes in both internal and external environments, simulate risk scenarios, and promptly adjust strategies for dynamic planning. This capability helps mitigate the likelihood of risks and minimize the potential damage they may cause. GAI can iterate on its algorithms, continuously refining its processes as it accumulates usage data and experience, thereby enhancing its adaptability. Furthermore, GAI can adjust the content it generates in response to evolving use cases, thus meeting users’ personalized needs. This adaptability allows GAI to dynamically adjust strategies based on external environmental changes and real-time feedback. By leveraging this capability, firms can enhance the flexibility and resilience of their supply chains, optimizing key processes such as inventory management, production, and logistics.
3 Framework based on GAI capabilities in SCM
In this section, we explore how GAI empowers SCM, addresses the challenges associated with its implementation, and presents potential solutions. The conceptual framework for SCM empowered by GAI is shown in Fig.2.
3.1 GAI enables SCM
GAI’s core capabilities offers significant advantages in data processing, decision support, and human-computer interaction within SCM. For instance, it efficiently extracts critical information, standardizes data formats, and summarizes large volumes of textual data, such as contracts and invoices. Moreover, GAI enhances decision-making by providing contextual insights, interpreting optimization results, and simulating various “what-if” scenarios. It further optimizes human-system interaction by integrating data sources and decision models, thus facilitating complex analyses and promoting agile, responsive operations. The combination of these capabilities enhances supply chain members’ visibility, accelerates decision-making, and fosters collaboration, ultimately strengthening the resilience and adaptability of supply chains. With ongoing advancements in GAI, its potential in SCM remains boundless. This subsection discusses how GAI drives improvements in SCM.
3.1.1 Empowering Forecasting: GAI Drives Precise Demand Insights
By leveraging deep learning and analyzing historical data, market trends, and external factors, GAI can develop more efficient and accurate forecasting models. The GAI’s ability to recognize complex demand patterns and provide data-driven insights enables enterprises to predict future demand with greater precision. In specific, GAI, with its learning, creativity, and perception and prediction capabilities, can integrate and analyze multimodal data, including historical sales data, natural language data, images, and videos. Furthermore, GAI can extract consumer sentiment and market feedback from social media, news, and other channels, thereby enabling accurate forecasting of potential demand fluctuations, as depicted in Fig.3. In practice, Cainiao introduces the GAI-powered big data model Tianjiπ, which leverages historical data, market trends, and potential disruptions to generate precise demand forecasts, thereby aiding businesses in optimizing inventory management and logistics resource allocation. By incorporating multidimensional data sources, such as consumer behavior, weather variations, and socio-economic indicators, Tianjiπ enhances the accuracy of demand forecasts, particularly under market volatility and supply chain uncertainties, thereby exhibiting superior adaptability.
3.1.2 Reshaping Procurement: GAI Leads Decision Optimization
GAI significantly enhances the accuracy of demand forecasting and optimizes the supplier selection process through its learning, creativity, and communication capabilities. As illustrated in Fig.4, GAI can leverage its perception and prediction abilities to integrate historical procurement data, inventory levels, and market trends to generate precise procurement demand forecasts. Based on these predictions, GAI helps formulate more scientific and effective purchasing plans. In supplier management, GAI can facilitate supplier negotiations and generate customized contract texts that align with procurement needs and legal regulations, thereby improving the efficiency of contract signing. Additionally, GAI can automatically generate comparative analysis reports by analyzing vast amounts of supplier data, enabling decision-makers to make prompt, accurate, and informed choices. In practice, Walmart implements the Pactum AI platform, which harnesses GAI to facilitate supplier interactions, automating negotiations and contract modifications. This approach not only generates substantial cost savings and optimized contract terms but also enhances supply chain efficiency by fostering more dynamic and data-driven decision-making. By leveraging GAI’s analytical capabilities to process vast data sets, Walmart can better synchronize its procurement strategies with market dynamics, thereby strengthening supplier relationships and optimizing operational performance.
3.1.3 Smart Inventory: GAI Optimizes Storage and Regulation
With its powerful capabilities in learning, creativity, collaboration, and interaction, GAI can more accurately predict inventory demand, identify market demand trends, and dynamically optimize replenishment and distribution strategies. This ensures that companies can meet production requirements and deliver products to customers in the shortest possible time. GAI intelligently optimizes inventory management processes by collecting and analyzing various financial and order data, as shown in Fig.5. Based on order dynamics, GAI efficiently schedules inbound and outbound operations of goods while optimizing warehouse layout based on the frequency of goods movements, ensuring optimal resource allocation and operational efficiency. Additionally, GAI can automatically generate accurate statistical reports based on goods movements, providing real-time support for decision-making and offering data-driven insights. In practice, ZARA develops its proprietary AI model to facilitate demand forecasting, sales analysis, and real-time inventory adjustments, resulting in improved inventory control, reduced surplus, and enhanced sales efficiency.
3.1.4 Revolutionizing Logistics: GAI Fuels Intelligent Operations
GAI in logistics management encompasses four key areas: intelligent scheduling and resource allocation, cross-department collaboration, logistics route optimization, and predictive maintenance, all of which significantly enhance management efficiency and flexibility, as illustrated in Fig.6. First, GAI integrates real-time data, such as orders, inventory, traffic, and weather, to enable intelligent scheduling and automatic resource allocation, optimizing the match between goods and vehicles, thereby reducing logistics costs and improving delivery efficiency. Second, GAI facilitates real-time cross-department collaboration by integrating supply chain information, such as shipper, warehouse, and receiver details, and leveraging automation systems to swiftly respond to unexpected changes, ensuring the efficient operation of logistics. Third, GAI analyzes multi-dimensional transportation network factors, thereby reducing delays and minimizing resource waste. Finally, GAI identifies potential equipment failures, such as those involving trucks or data storage devices, proactively schedules maintenance tasks, and reduces failure rates, ultimately enhancing the reliability and continuity of logistics systems. In practice, Baidu Maps introduces its GAI-powered Logistics Big Model Beta, which is utilized for logistics address parsing and dispatch decision-making. By employing advanced AI algorithms, this model enhances both the accuracy and speed of address identification while optimizing route planning and dispatch decision-making, thereby substantially improving logistics efficiency.
3.1.5 Forward-Looking Risk Management: GAI Builds Resilience
Through intelligent analysis and predictive modeling, GAI not only assists companies in identifying potential risks and optimizing decision-making processes but also offers flexible countermeasures. As depicted in Fig.7, GAI analyzes historical data, market trends, and external environmental changes to predict potential supply chain disruptions, enabling companies to take proactive measures in advance. In the event that risks have already materialized, GAI can rapidly identify and assess their severity, coordinating all systems to respond swiftly and ensuring that the supply chain returns to normal operations. Furthermore, GAI enhances supply chain resilience and responsiveness by automating responses and dynamically adjusting strategies, effectively addressing sudden risk events and ensuring stable supply chain operations. In practice, IBM utilizes GAI models to predict and respond to supplier disruptions, natural disasters, and geopolitical events, thereby ensuring operational stability and continuity.
In short, compared to traditional SCM, the GAI-enabled SCM has made notable advancements in key areas, including demand forecasting, procurement management, inventory management, logistics management and risk response. Tab.2 presents a comparison of these two types of SCM across these key aspects.
3.2 Challenges and solutions
Despite the significant potential applications of GAI in SCM, its widespread deployment and practical implementation face numerous challenges. These challenges not only involve technical complexities but also encompass multidimensional constraints related to data, ethics, society, and cost. Therefore, this section provides an in-depth analysis of the potential risks and practical issues associated with the use of GAI in SCM, along with proposed solutions to address these challenges, as illustrated in Tab.3.
4 Frontier research agenda
As the body of literature on GAI’s applications in SCM continues to grow, our understanding of its transformative potential remains limited. To address the pressing questions surrounding this emerging field, this study identifies four frontier research directions, categorized into technology-driven approaches and management innovation practices, as shown in Tab.4.
In the domain of technology-driven directions, the application of GAI in smart supply chain design and risk prediction offers significant opportunities. Research in this area focuses on leveraging GAI to design supply chain networks that can adapt to complex environments and high-risk scenarios. The key challenges include utilizing GAI to predict potential supply chain risks, developing adaptive response strategies, and optimizing supply chain structures to enhance resilience and risk mitigation. Furthermore, the intersection of GAI with emerging technologies holds the potential to further enhance predictive accuracy and decision-making processes. By integrating advanced GAI techniques with innovative technologies, this field can create more robust, adaptive, and risk-aware supply chains, better equipped to address future challenges.
In the realm of management innovation practices, the ethical and social implications of GAI in supply chain decision-making represent crucial research areas. These challenges encompass issues such as data bias, decision transparency, and the potential risk of job displacement. Central to these discussions are the design of interpretable AI systems to improve transparency and strategies for mitigating the impact of GAI on workforce transformations within supply chains. Moreover, GAI’s role in sustainable supply chain design is gaining attention, particularly in optimizing carbon emissions, resource recycling, and low-carbon logistics. Research in this area aims to balance economic efficiency with environmental sustainability, supporting businesses in achieving carbon neutrality and long-term sustainable development.
5 Conclusions
Supply chain operations have become increasingly complex over the years, with businesses facing growing pressure to deliver goods more rapidly and efficiently while simultaneously reducing costs. GAI offers numerous advantages for SCM. Despite the abundance of literature on the role of GAI in enhancing supply chain performance, a comprehensive theoretical framework for the construction of GAI applications and their empowerment mechanisms within SCM remains underdeveloped. This study presents a critical analysis and a theoretical framework for SCM enabled by GAI. In specific, we first identify the core GAI capabilities necessary for constructing the SCM framework. We then examine the empowerment mechanisms and challenges associated with GAI in SCM, offering corresponding solutions. Subsequently, we identify key gaps and propose a comprehensive research agenda, emphasizing the SCM framework empowered by GAI. The insights and recommendations provided in this study aim to assist firms in building flexible, robust, and sustainable supply chains in the era of GAI.
PDF
(1932KB)
