The Ripple Effect of Vulnerabilities in Maven Central: Prevalence, Propagation, and Mitigation Challenges
Ehtisham Ul Haq, Song Wang, Robert S. Allison · 2025
Read The Ripple Effect of Vulnerabilities in Maven Central: Prevalence, Propagation, and Mitigation Challenges by Ehtisham Ul Haq, Song Wang, Robert S. Allison (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-04175.txt.
Introduction
Ehtisham Ul Haq, Song Wang, and Robert Allison trace how security flaws ripple through Maven Central, the main library hub for Java software. Package managers speed up development, but each borrowed library can carry flaws into every project that depends on it. The dependence can be direct, or it can run through chains of further libraries. The team measured both kinds of exposure using public vulnerability records.
In a sample of around four million releases, only about one percent carried a direct vulnerability. Yet roughly 46.8 percent were affected once flaws in their dependency chains were counted. The paper then turns to why fixing this is hard. Updates must move through the same chains that spread the flaws, and maintainers often lack a clear view of what they indirectly rely on.
Catalog record
| CREATOR | Ehtisham Ul Haq, Song Wang, Robert S. Allison |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 17 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-04175
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The Ripple Effect of Vulnerabilities in Maven Central: Prevalence, Propagation, and Mitigation Challenges Ehtisham Ul Haq, Song Wang, Robert S. Allison Abstract The widespread use of package managers like Maven has accelerated software development but has also introduced significant security risks due to vulnerabilities in dependencies. In this study, we analyze the prevalence and impact of vulnerabilities within the Maven Central ecosystem, using Common Vulnerabilities and Exposures (CVE) data from OSV.dev and a subsample enriched with aggregated CVE data (CVE_AGGREGATED), which captures both direct and transitive vulnerabilities. In our subsample of around 4 million releases, we found that while only about 1% of releases have direct vulnerabilities, approximately 46.8% are affected by transitive vulnerabilities. This highlights how a small number of vulnerable yet influential artifacts can impact a vast portion of the ecosystem. Moreover, our analysis shows that vulnerabilities propagate rapidly through dependency networks and that more central artifacts (those with a high number of dependents) are not necessarily less vulnerable. We also observed that the time taken to patch vulnerabilities, including those of high or critical severity, often spans several years. Additionally, we found that dependents of artifacts tend to prefer presumably non-vulnerable versions; however, some continue to use vulnerable versions, indicating challenges in adopting patched releases. These findings highlight the critical need for improved dependency management practices and timely vulnerability remediation to enhance the security of software ecosystems. I Introduction Extensive code reuse through package managers has become integral to modern software development, enabling rapid application building by leveraging existing libraries. Maven, a prominent package manager for Java, simplifies dependency management by automating the inclusion of both direct and transitive dependencies. However, this convenience introduces significant security risks. Vulnerabilities in dependencies can propagate through complex and deep dependency chains, potentially affecting a vast number of projects [ 1 , 2 ] . The problem is exacerbated by the fact that developers often lack visibility into the vulnerabilities present in their transitive dependencies. Studies have shown that a considerable number of software projects inadvertently include vulnerable libraries [ 3 ] . Although not all vulnerabilities necessarily impact dependent projects, as their exploitability depends on code usage, those that do have led to significant security breaches and data leaks. For instance, the Equifax data breach in 2017, which exposed the personal information of millions of individuals, was attributed to an unpatched vulnerability in a third-party library [ 4 ] . This incident highlights the importance of timely vulnerability management in dependencies. Despite awareness of these risks, many developers struggle to keep their dependencies updated. Kula et al. found that outdated libraries with known vulnerabilities are prevalent in software projects, mainly due to challenges in dependency management and the fear of breaking changes [ 5 ] . Addressing these issues is essential for improving software security. By analyzing the prevalence and propagation of vulnerabilities, we can identify areas for improvement and develop strategies to mitigate risks associated with dependency management. In this paper, we aim to fill the gap in understanding how vulnerabilities affect the Maven Central ecosystem. Specifically, we address the following research questions: RQ1: ( Distribution ) What proportion of releases have known vulnerabilities? What is the proportion of releases directly and transitively impacted? RQ2: ( Propagation ) How do vulnerabilities propagate through the dependency network, and which projects are most affected? RQ3: ( Lifetime ) What is the average time taken to patch a vulnerability in a dependency? RQ4: ( Response ) How do users of an artifact react to the discovery of a vulnerability in that artifact? To address these questions, we analyzed a dataset from the Maven Central repository, enriched with Common Vulnerabilities and Exposures (CVE) data sourced from OSV.dev [ 6 ] . Using Neo4j and Cypher queries, we conducted experiments to uncover empirical insights into the prevalence and propagation of vulnerabilities, the timeliness of patches, and user behavior in response to vulnerabilities. II Experiments and Results The experiments utilized the latest dataset snapshot from August 30th, 2024, comprising 658,078 artifact nodes, 14,459,139 release nodes, and 134,119,545 dependency edges. The database was enriched with CVE data up to September 4, 2024. We added additional CVE_AGGREGATED values (capturing both direct and transitive vulnerabilities) using version 2.1.0 of the Goblin Weaver tool [ 7 ] . Due to the dataset’s extensive size and the computational intensity of the enrichment process, the CVE_AGGREGATED data enrichment was limited to a subset of 4,095,768 releases. II-A RQ1: Distribution of Vulnerabilities Experiment 1 We examined the subsample enriched with CVE_AGGREGATED data to analyze the prevalence of vulnerabilities. Using Cypher queries on JSON-formatted CVE data, we identified releases with at least one direct vulnerability (non-empty CVE field) and those with only transitive vulnerabilities (non-empty CVE_AGGREGATED but empty CVE field). Result: We found that 40,809 releases (1% of the subset) had at least one direct vulnerability and 1,916,314 releases (46.8%) had at least one transitive vulnerability. Figure 1 illustrates the proportion of releases affected by vulnerabilities. Fig. 1 : Proportion of Vulnerable Releases (from 4,095,768 Releases with Enriched CVE_Aggregated Data) Finding 1: These results show that while direct vulnerabilities impact a very small fraction of releases, transitive vulnerabilities affect nearly half. It underscores the importance of considering both direct and transitive dependencies when assessing a project’s security posture. II-B RQ2: Propagation of Vulnerabilities Experiment 2.1 To analyze how vulnerabilities propagate through the dependency network, we examined artifacts with direct vulnerabilities: • From the complete dataset (around 14 million releases), we identified 77,393 releases with at least one direct vulnerability. • These releases were linked to 1,411 distinct artifacts, which we labeled as ArtifactsWithDirectVulnerabilities . • Using Cypher queries, we calculated the number of dependent releases at: – Depth 1: Direct dependents of vulnerable artifacts. – Depth 2: Dependents of direct dependents at depth 1. Result: The average dependents per artifact were 17,993 at depth 1 and 142,948 at depth 2. Although 292 artifacts had no direct dependents, the remaining 1,119 showed significantly higher numbers at depth 2, as illustrated in Figure 2 . Fig. 2 : Depth 2 dependents vs. depth 1 dependents of ArtifactsWithDirectVulnerabilities on a logarithmic scale Finding 2: The exponential growth in the number of dependents from depth 1 to depth 2 highlights the cascading effect of vulnerabilities in the dependency network. This vast amplification of impact underscores the need for strategies to manage and contain vulnerabilities at their source, as their unchecked propagation can affect a disproportionately large number of releases. Experiment 2.2 To examine whether central artifacts (those with many dependents) are less vulnerable, we analyzed artifacts with aggregated vulnerabilities of moderate to critical severity. This included 1,953,496 releases from 145,174 artifacts labeled as ArtifactsWithSignificantVulnerabilities . We calculated unique vulnerability counts to avoid over-penalizing artifacts for unpatched versions. The artifacts were grouped into four centrality buckets ( Low , Medium , High , Very High ) based on the number of dependents and the average vulnerabilities per bucket were calculated. Result: The analysis revealed a subtle increase in vulnerabilities with centrality. Statistical tests confirmed significant differences across centrality buckets (ANOVA: F=4.26, p=0.005; Kruskal-Wallis: H=109.92, p < 1e-22), with post-hoc analysis showing more vulnerabilities in the ’Very High’ bucket compared to the ’Low’ bucket (Tukey HSD: p=0.002). Figure 3