Open Source, Open Threats? Investigating Security Challenges in Open-Source Software
Seyed Ali Akhavani *, Behzad Ousat *, Amin Kharraz, * Authors contributed equally to this work. · 2025
Read Open Source, Open Threats? Investigating Security Challenges in Open-Source Software by Seyed Ali Akhavani *, Behzad Ousat *, Amin Kharraz, * Authors contributed equally to this work. (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-2506-12995.txt.
Introduction
Seyed Ali Akhavani, Behzad Ousat, and Amin Kharraz wrote this paper in two thousand twenty-five. It probes the safety weak spots of open-source software, which is now used in nearly every field but whose open growth also opens doors to attack.
The authors study tens of thousands of flaw reports from GitHub's advisory base and Snyk, across many packages and ten coding languages. Reports are rising far faster than the stock of packages, and flaws now live longer before they are fixed. A small set of weakness types accounts for over half of all reports. Most alarming, near half the reports in the NPM store are tied to packages made with intent to harm. The paper ends with a close look at how those bad packages work.
Catalog record
| CREATOR | Seyed Ali Akhavani *, Behzad Ousat *, Amin Kharraz, * Authors contributed equally to this work. |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 63 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-12995
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Open Source, Open Threats? Investigating Security Challenges in Open-Source Software Seyed Ali Akhavani *, Behzad Ousat *, Amin Kharraz, * Authors contributed equally to this work. Abstract Open-source software (OSS) has become increasingly more popular across different domains. However, this rapid development and widespread adoption come with a security cost. The growing complexity and openness of OSS ecosystems have led to increased exposure to vulnerabilities and attack surfaces. This paper investigates the trends and patterns of reported vulnerabilities within OSS platforms, focusing on the implications of these findings for security practices. To understand the dynamics of OSS vulnerabilities, we analyze a comprehensive dataset comprising 31,267 unique vulnerability reports from GitHub’s advisory database and Snyk.io, belonging to 14,675 packages across 10 programming languages. Our analysis reveals a significant surge in reported vulnerabilities, increasing at an annual rate of 98%—far outpacing the 25% average annual growth in the number of open-source software (OSS) packages. Additionally, we observe an 85% increase in the average lifespan of vulnerabilities across ecosystems during the studied period, indicating a potential decline in security. We identify the most prevalent Common Weakness Enumerations (CWEs) across programming languages and find that, on average, just seven CWEs are responsible for over 50% of all reported vulnerabilities. We further examine these commonly observed CWEs and highlight ecosystem-specific trends. Notably, we find that vulnerabilities associated with intentionally malicious packages comprise 49% of reports in the NPM ecosystem and 14% in PyPI—an alarming indication of targeted attacks within package repositories. We conclude with an in-depth discussion of the characteristics and attack vectors associated with these malicious packages. I Introduction The open-source software (OSS) landscape has seen exponential growth over the past decade, becoming a backbone for a wide range of industries, including aerospace [ 1 ] , energy systems [ 2 , 3 ] , finance [ 4 , 5 ] , healthcare [ 6 ] , and government projects. As the adoption of OSS starts to dominate across different sectors, its inherent benefits—such as transparency, collaboration, and rapid innovation—are accompanied by important challenges. The open and decentralized nature of OSS, while promoting collaboration and innovation, also creates opportunities for adversaries to exploit vulnerabilities [ 7 ] , compromise supply chains [ 8 , 9 ] , and introduce malicious code into widely-used libraries [ 10 , 11 , 12 ] . Government entities, policymakers, and regulators recognize the need for stronger security measures in the open-source environment. The recent United States Securing Open Source Software Act represents a significant step toward mandating security improvements and creating frameworks for the protection of open-source libraries and dependencies [ 13 , 14 ] . Similarly, United States’ Cybersecurity and Infrastructure Security Agency (CISA) has published the Open Source Software Security Roadmap , outlining its strategy to ensure a secure open-source ecosystem [ 15 ] . These initiatives reflect growing concerns over the reliance on OSS in critical infrastructure, highlighting challenges like dependency visibility, large-scale vulnerability management, and human errors in maintaining packages. Our work is guided by three primary research questions. 1) How have vulnerabilities evolved in different ecosystems?, 2) How long do vulnerabilities persist and spread across packages?, 3) What are the unique vulnerability patterns in different ecosystems? To answer these questions, we generated a dataset of vulnerability reports from 2017 to January 2025. The data consists of over 31,267 reported vulnerabilities from two advisory databases–GitHub Advisory Database and Snyk.io– that cover 14,675 open-source repositories publicly available on GitHub. We also have reports of 4,456 malicious packages whose source code is no longer available since they have been permanently removed. We examine vulnerabilities across 10 programming languages –C, C++, PHP, Rust, JavaScript (Node.JS), Go, Java, .Net, Python, Ruby–, and eight distinct package managers –PyPi, NPM, Composer, Crates, Go, Maven, NuGet, and RubyGems– which are foundational to software and web development [ 16 , 17 , 18 , 19 ] . This longitudinal analysis allows us to track the growth and distribution of vulnerabilities across programming languages. In the following, we highlight some of the major findings of this paper. The rapid adoption of open-source platforms is being overshadowed by a faster rise in reported vulnerabilities. The analysis reveals a substantial increase in the number of packages across open-source platforms, with an average annual growth rate of 25.16%. However, this growth is accompanied by an alarming rise in reported vulnerable packages, which have increased at a much faster rate—98% annually. This suggests that as package managers rapidly expand their number of packages, the security risks associated with them are growing disproportionately. For example, in 2022, there were 2,490 vulnerabilities reported for Maven and 1,593 for Composer. The sharp increase in vulnerabilities, particularly in high-growth platforms, highlights a critical security gap, as the rate of vulnerabilities is outpacing the platform’s expansion, making open-source software an increasingly attractive target for attackers. The expanding scale of open-source ecosystems does not necessarily translate into improved security. We analyze vulnerabilities at the package level and find that the time a vulnerability remains in the ecosystem before being fixed has increased by 95% from 2017 to 2024. Additionally, we examine vulnerability concentration patterns to assess whether security issues stem from a small number of packages or are dispersed across the ecosystem. Our analysis reveals distinct trends—some platforms exhibit a high concentration of vulnerabilities in key packages, while others show a broader, more distributed spread. For instance, Composer (4.92 vulnerabilities per vulnerable package) and C/C++ (3.96) have vulnerabilities concentrated in a few critical packages. In contrast, ecosystems such as NPM (1.75 vulnerabilities per package) and Go (1.70) show a more even distribution of vulnerabilities. The evaluation of CWEs across ecosystems reveals distinct vulnerability patterns shaped by both common threats and language-specific characteristics. Our results reveal key trends in vulnerability distribution, highlighting the prevalence of certain vulnerabilities and the impact of language-specific characteristics. Our data demonstrates that, on average of the investigated platforms, only seven CWEs were responsible for over 50% of the vulnerability reports. Notably, CWE-79 (Cross-Site Scripting) and CWE-22 (Path Traversal) pose a major threat in ecosystems like Composer, NPM, PyPI, and Maven, reflecting widespread security challenges in web applications. On the other hand, some CWEs are deeply tied to language characteristics, such as CWE-122 (Heap-based Buffer Overflow) in C/C++ or CWE-1321 (Prototype Pollution) in JavaScript. We specifically focus on CWE-506 (Embedded Malicious Code) that has emerged as a significant issue in platforms like NPM and PyPI, where 48.58% and 13.92% of all vulnerabilities, represent intentional attacks underscoring the growing concern over supply chain attacks. We hope this work to raise awareness about the increasing need for systematic approaches to enhance visibility in open-source software. We also hope that the insights provided in this paper will inspire new directions in developing automated tools that can effectively model vulnerabilities and augment developers’ experience to build more robust code and test cases. We summarize our contributions as follows: • We present a longitudinal study that explores open-source software package across 10 programming languages and eight distinct package managers from 2017 to 2025, measuring the growth rate of packages and vulnerabilities. • We investigate the lifespan of vulnerabilities over the last years across the ecosystems and discuss the uneven distribution of vulnerabilities across packages, identifying