CVE-2024-12055
Published: 20 March 2025
Summary
CVE-2024-12055 is a high-severity Out-of-bounds Read (CWE-125) vulnerability in Ollama Ollama. Its CVSS base score is 7.5 (High).
Operationally, exploitation aligns with the MITRE ATT&CK technique Application or System Exploitation (T1499.004); ranked at the 21.8th 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 NLP and Transformers; in the Data-Related Vulnerabilities risk domain.
The strongest mitigations our analysis identified are NIST 800-53 SC-5 (Denial-of-service Protection) and SI-10 (Information Input Validation).
Deeper analysis
CVE-2024-12055 is a vulnerability in Ollama versions up to and including 0.3.14, stemming from an out-of-bounds read in the gguf.go file and classified under CWE-125. This flaw allows a malicious user to create a customized GGUF model file that can be uploaded to a public Ollama server. When the server processes the malicious model, it crashes, resulting in a Denial of Service (DoS) condition. The issue carries a CVSS v3.1 base score of 7.5 (AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H), highlighting its high availability impact.
Any unauthenticated attacker with network access can exploit this vulnerability due to its lack of prerequisite privileges or user interaction. By crafting and uploading the malicious GGUF model file to the public Ollama server, the attacker triggers the out-of-bounds read during model processing, causing the server to crash and disrupting service availability for all users.
For mitigation details, security practitioners should consult the advisory referenced at https://huntr.com/bounties/7b111d55-8215-4727-8807-c5ed4cf1bfbe, which documents the issue discovered through a bug bounty program. Upgrading to Ollama versions beyond 0.3.14 is implied as a resolution based on the affected version scope.
EU & UK References
- 🇪🇺 ENISA EUVD: EUVD-2025-7036
Vulnerability details
A vulnerability in Ollama versions <=0.3.14 allows a malicious user to create a customized gguf model file that can be uploaded to the public Ollama server. When the server processes this malicious model, it crashes, leading to a Denial of…
more
Service (DoS) attack. The root cause of the issue is an out-of-bounds read in the gguf.go file.
- CWE(s)
AI Security AnalysisAI
- AI Category
- NLP and Transformers
- Risk Domain
- Data-Related Vulnerabilities
- OWASP Top 10 for LLMs 2025
- None mapped
- Classification Reason
- Matched keywords: ollama
Related Threats
MITRE ATT&CK Enterprise TechniquesAI
Why these techniques?
The vulnerability enables a denial of service by uploading a malicious gguf model file to the public Ollama server, causing an out-of-bounds read crash during processing, matching application exploitation for endpoint DoS.
CVEs Like This One
Affected Assets
Mitigating Controls
Mitigating Controls (NIST 800-53 r5) AI
Remediating the out-of-bounds read flaw in gguf.go by patching to Ollama versions beyond 0.3.14 directly eliminates the vulnerability causing the server crash.
Validating uploaded GGUF model files for proper structure and bounds before processing prevents malicious inputs from triggering the out-of-bounds read.
Implementing denial-of-service protections mitigates server crashes from processing malicious GGUF files on public Ollama servers.