Understanding Security Risks of AI Agents' Dependency Updates

• URL: http://arxiv.org/abs/2601.00205v1
• Date: Thu, 01 Jan 2026 04:44:18 GMT
• Title: Understanding Security Risks of AI Agents' Dependency Updates
• Abstract summary: Dependency changes can substantially alter a project's security posture.<n>We study 117,062 dependency changes from agent- and human-authored pull requests across seven ecosystems.<n>Agent-driven dependency work yields a net vulnerability increase of 98, whereas human-authored work yields a net reduction of 1,316.
• Score: 8.978334182646465
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Package dependencies are a critical control point in modern software supply chains. Dependency changes can substantially alter a project's security posture. As AI coding agents increasingly modify software via pull requests, it is unclear whether their dependency decisions introduce distinct security risks. We study 117,062 dependency changes from agent- and human-authored pull requests across seven ecosystems. Agents select known-vulnerable versions more often than humans (2.46% vs. 1.64%), and their vulnerable selections are more disruptive to remediate, with 36.8% requiring major-version upgrades compared to 12.9% for humans, despite patched alternatives existing in most cases. At the aggregate level, agent-driven dependency work yields a net vulnerability increase of 98, whereas human-authored work yields a net reduction of 1,316. These findings motivate pull-request-time vulnerability screening and registry-aware guardrails to make agent-driven dependency updates safer.

Related papers

• HazardAuditor: From Executable Threats to Safer Computer-Use Agents [50.06351661912873]
  Existing guard models target static prompts and responses.<n>Existing safety platforms produce evaluation verdicts rather than the normalized supervision a guard model needs to learn.<n>We introduce HazardAuditor, an execution-grounded framework that closes both gaps.
  arXiv  Detail & Related papers  (2026-09-14T07:09:33Z)
• EvoSafeHarness: Evolving Model- and Domain-Specific Harnesses for Securing Agents [74.54929751174289]
  EvoSafeHarness is a safety-specific optimization framework that synthesizes a deployable harness for a frozen model in a target domain.<n>It reduces average attack success rate from 45.6% to 10.0% at a 3.3-point utility cost.<n>It also achieves the best score on Agent-SafetyBench for every victim.
  arXiv  Detail & Related papers  (2026-09-05T05:56:41Z)
• Trust but Verify? Uncovering the Security Debt of Autonomous Coding Agents [3.30025498453752]
  This paper presents a large-scale empirical study using the AIDev dataset to characterize security code smells in agent-generated pull requests.<n>We find that 38.9% of agent-generated PRs contain at least one security smell, with supply chain integrity issues accounting for 82.3% of all detected security smells.<n>Human collaborators are responsible for introducing 67.6% of genuine leaked secrets within these agent-assisted systems.
  arXiv  Detail & Related papers  (2026-07-14T06:59:41Z)
• Lessons from Penetration Tests on Large-Scale Agent Systems [7.61892296424389]
  Execution-capable AI agents interact extensively with multiple layers of the computing stack.<n>It remains unclear whether proprietary agent systems exhibit similar security weaknesses.<n>We present findings from two penetration tests conducted in 2025 against proprietary agent products.
  arXiv  Detail & Related papers  (2026-05-26T13:58:27Z)
• On-Policy Self-Evolution via Failure Trajectories for Agentic Safety Alignment [54.30690671490447]
  Existing safety-alignment signals are largely response-level or off-policy.<n>FATE transforms verifier-scored failures into repair supervision without expert demonstrations.<n>FATE reduces attack success rate by 33.5%, harmful compliance by 82.6%, and improves external trajectory-safety diagnosis by 6.5%.
  arXiv  Detail & Related papers  (2026-05-12T09:56:28Z)
• DecodingTrust-Agent Platform (DTap): A Controllable and Interactive Red-Teaming Platform for AI Agents [121.77550256034]
  DecodingTrust-Agent Platform (DTap) is a controllable and interactive red-teaming platform for AI agents.<n>DTap-Red is the first autonomous red-teaming agent that explores diverse injection vectors and autonomously discovers effective attack strategies.<n>Through DTap, we conduct large-scale evaluations of popular AI agents built on various backbone models.
  arXiv  Detail & Related papers  (2026-05-06T11:59:48Z)
• Trust as Monitoring: Evolutionary Dynamics of User Trust and AI Developer Behaviour [59.27481013639016]
  We study how user trust strategies and developer choices between safe and unsafe AI co-evolve under different levels of monitoring cost and institutional regimes.<n>We find three robust long-run regimes: no adoption with unsafe development, unsafe but widely adopted systems, and safe systems that are widely adopted.
  arXiv  Detail & Related papers  (2026-03-25T19:04:23Z)
• OMNI-LEAK: Orchestrator Multi-Agent Network Induced Data Leakage [59.3826294523924]
  We investigate the security vulnerabilities of a popular multi-agent pattern known as the orchestrator setup.<n>We report the susceptibility of frontier models to different categories of attacks, finding that both reasoning and non-reasoning models are vulnerable.
  arXiv  Detail & Related papers  (2026-02-13T21:32:32Z)
• Security in the Age of AI Teammates: An Empirical Study of Agentic Pull Requests on GitHub [4.409447722044799]
  This study aims to characterize how autonomous coding agents contribute to software security in practice.<n>We conduct a large-scale empirical analysis of agent-authored PRs using the AIDev dataset.<n>We then analyze prevalence, acceptance outcomes, and review latency across autonomous agents, programming ecosystems, and types of code changes.
  arXiv  Detail & Related papers  (2026-01-01T21:14:11Z)
• Breaking Agent Backbones: Evaluating the Security of Backbone LLMs in AI Agents [36.2255033141489]
  AI agents powered by large language models (LLMs) are being deployed at scale, yet we lack a systematic understanding of how the choice of backbone LLM affects agent security.<n>We introduce threat snapshots: a framework that isolates specific states in an agent's execution flow where vulnerabilities manifest.<n>We apply this framework to construct the $operatornameb3$ benchmark, a security benchmark based on 194331 unique crowdsourced adversarial attacks.
  arXiv  Detail & Related papers  (2025-10-26T10:36:42Z)
• AdvEvo-MARL: Shaping Internalized Safety through Adversarial Co-Evolution in Multi-Agent Reinforcement Learning [78.5751183537704]
  AdvEvo-MARL is a co-evolutionary multi-agent reinforcement learning framework that internalizes safety into task agents.<n>Rather than relying on external guards, AdvEvo-MARL jointly optimize attackers and defenders.
  arXiv  Detail & Related papers  (2025-10-02T02:06:30Z)
• Security Challenges in AI Agent Deployment: Insights from a Large Scale Public Competition [101.86739402748995]
  We run the largest public red-teaming competition to date, targeting 22 frontier AI agents across 44 realistic deployment scenarios.<n>We build the Agent Red Teaming benchmark and evaluate it across 19 state-of-the-art models.<n>Our findings highlight critical and persistent vulnerabilities in today's AI agents.
  arXiv  Detail & Related papers  (2025-07-28T05:13:04Z)
• WebGuard: Building a Generalizable Guardrail for Web Agents [59.31116061613742]
  WebGuard is the first dataset designed to support the assessment of web agent action risks.<n>It contains 4,939 human-annotated actions from 193 websites across 22 diverse domains.
  arXiv  Detail & Related papers  (2025-07-18T18:06:27Z)
• SafeMobile: Chain-level Jailbreak Detection and Automated Evaluation for Multimodal Mobile Agents [58.21223208538351]
  This work explores the security issues surrounding mobile multimodal agents.<n>It attempts to construct a risk discrimination mechanism by incorporating behavioral sequence information.<n>It also designs an automated assisted assessment scheme based on a large language model.
  arXiv  Detail & Related papers  (2025-07-01T15:10:00Z)
• SafeAgent: Safeguarding LLM Agents via an Automated Risk Simulator [77.86600052899156]
  Large Language Model (LLM)-based agents are increasingly deployed in real-world applications.<n>We propose AutoSafe, the first framework that systematically enhances agent safety through fully automated synthetic data generation.<n>We show that AutoSafe boosts safety scores by 45% on average and achieves a 28.91% improvement on real-world tasks.
  arXiv  Detail & Related papers  (2025-05-23T10:56:06Z)
• Information Retrieval Induced Safety Degradation in AI Agents [52.15553901577888]
  This study investigates how expanding retrieval access affects model reliability, bias propagation, and harmful content generation.<n>Retrieval-enabled agents built on aligned LLMs often behave more unsafely than uncensored models without retrieval.<n>These findings underscore the need for robust mitigation strategies to ensure fairness and reliability in retrieval-enabled and increasingly autonomous AI systems.
  arXiv  Detail & Related papers  (2025-05-20T11:21:40Z)
• AgentVigil: Generic Black-Box Red-teaming for Indirect Prompt Injection against LLM Agents [54.29555239363013]
  We propose a generic black-box fuzzing framework, AgentVigil, to automatically discover and exploit indirect prompt injection vulnerabilities.<n>We evaluate AgentVigil on two public benchmarks, AgentDojo and VWA-adv, where it achieves 71% and 70% success rates against agents based on o3-mini and GPT-4o.<n>We apply our attacks in real-world environments, successfully misleading agents to navigate to arbitrary URLs, including malicious sites.
  arXiv  Detail & Related papers  (2025-05-09T07:40:17Z)
• The Ripple Effect of Vulnerabilities in Maven Central: Prevalence, Propagation, and Mitigation Challenges [8.955037553566774]
  We analyze the prevalence and impact of vulnerabilities within the Maven Central ecosystem using Common Vulnerabilities and Exposures data.<n>In our subsample of around 4 million releases, we found that while only about 1% of releases have direct vulnerabilities.<n>We also observed that the time taken to patch vulnerabilities, including those of high or critical severity, often spans several years.
  arXiv  Detail & Related papers  (2025-04-05T13:45:27Z)
• Pinning Is Futile: You Need More Than Local Dependency Versioning to Defend against Supply Chain Attacks [23.756533975349985]
  Recent high-profile incidents in open-source software have raised practitioner attention on software supply chain attacks.<n>Security practitioners advocate pinning dependency to specific versions rather than floating in version ranges.<n>We quantify, through counterfactual analysis and simulations, the security and maintenance impact of version constraints in the npm ecosystem.
  arXiv  Detail & Related papers  (2025-02-10T16:50:48Z)
• Dependency Practices for Vulnerability Mitigation [4.710141711181836]
  We analyze more than 450 vulnerabilities in the npm ecosystem to understand why dependent packages remain vulnerable. We identify over 200,000 npm packages that are infected through their dependencies. We use 9 features to build a prediction model that identifies packages that quickly adopt the vulnerability fix and prevent further propagation of vulnerabilities.
  arXiv  Detail & Related papers  (2023-10-11T19:48:46Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.