CVE-2024-4881
Published: 06 June 2024
Summary
CVE-2024-4881 is a high-severity Absolute Path Traversal (CWE-36) vulnerability in Lollms Lollms. Its CVSS base score is 7.5 (High).
Operationally, exploitation aligns with the MITRE ATT&CK technique Exploit Public-Facing Application (T1190); ranked at the 43.8th percentile by exploit likelihood (below the median); it is not currently listed in the CISA KEV catalog; a public proof-of-concept is referenced.
This vulnerability is AI-related — categorised as Enterprise AI Assistants; in the Privacy and Disclosure risk domain; MITRE ATLAS techniques in scope: Adversarial AI Attack Implementations (AML.T0016.000), Invert AI Model (AML.T0024.001), Financial Harm (AML.T0048.000).
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2024-0095
Vulnerability details
A path traversal vulnerability exists in the parisneo/lollms application, affecting version 9.4.0 and potentially earlier versions, but fixed in version 5.9.0. The vulnerability arises due to improper validation of file paths between Windows and Linux environments, allowing attackers to traverse…
more
beyond the intended directory and read any file on the Windows system. Specifically, the application fails to adequately sanitize file paths containing backslashes (`\`), which can be exploited to access the root directory and read, or even delete, sensitive files. This issue was discovered in the context of the `/user_infos` endpoint, where a crafted request using backslashes to reference a file (e.g., `\windows\win.ini`) could result in unauthorized file access. The impact of this vulnerability includes the potential for attackers to access sensitive information such as environment variables, database files, and configuration files, which could lead to further compromise of the system.
- CWE(s)
AI Security AnalysisAI
- AI Category
- Enterprise AI Assistants
- Risk Domain
- Privacy and Disclosure
- OWASP Top 10 for LLMs 2025
- None mapped
- Classification Reason
- parisneo/lollms is an AI assistant platform (Lord of Large Language Models WebUI) for interacting with LLMs, fitting the Enterprise AI Assistants category as a deployed application for AI model usage and management.
Related Threats
MITRE ATT&CK Enterprise TechniquesAI
Why these techniques?
Path traversal in public-facing web app (/user_infos endpoint) enables exploitation (T1190), arbitrary file reads for data collection and discovery (T1005, T1083), and potential deletions (T1070.004).
MITRE ATLAS TechniquesAI
MITRE ATLAS techniques
Affected Assets
Mitigating Controls
Likely Mitigating Controls AI
Per-CVE control mapping for this CVE has not run yet; the list below is derived from the weakness types (CWEs) cited in the NVD entry.
Validates pathnames and filenames to prevent traversal outside intended directories.