Submodular functions have been a powerful mathematical model for a wide range of real-world applications. Recently, submodular functions are becoming increasingly important in machine learning (ML) for modelling notions such as information and redundancy among entities such as data and features. Among these applications, a key question is payoff allocation, i.e., how to evaluate the importance of each entity towards a collective objective? To this end, classic solution concepts from cooperative game theory offer principled approaches to payoff allocation. However, despite the extensive body of gametheoretic literature, payoff allocation in submodular games is relatively under-researched. In particular, an important notion that arises in the emerging submodular applications is redundancy, which may occur from various sources such as abundant data or malicious manipulations where a player replicates its resource and acts under multiple identities. Though many gametheoretic solution concepts can be directly used in submodular games, naively applying them for payoff allocation in these settings may incur robustness issues against replication. In this paper, we systematically study the replication manipulation in submodular games and investigate replication robustness, a metric that quantitatively measures the robustness of solution concepts against replication. Using this metric, we present conditions which theoretically characterise the robustness of semivalues, a wide family of solution concepts including the Shapley and Banzhaf value. Moreover, we empirically validate our theoretical results on an emerging submodular ML application-ML data markets.Impact Statement-With the increasing take-up of ML techniques in real-world settings, payoff allocation has significant impact towards fairness, trustworthiness, safety, and knowledge discovery in ML applications, e.g., performing analysis or debugging of ML systems by finding the key contributors or bottleneck entities. Many emerging ML applications exhibit submodular characteristics, while properties of classic game-theoretic payoff allocation on submodular games are under-researched. This paper investigats an important issue of redundancy arising from replication in submodular ML applications. Using a replication robustness metric, we provide theoretical guarantees for the robustness of common game-theoretic payoff allocation methods against replication. Our findings can guide the use of gametheoretic payoff allocation in submodular ML applications, and impact real-world applications and future research on payoff allocation in ML systems in general, such as fair compensation in multi-party ML systems and feature importance interpretation in the medical domains.