CVE-2026-85673: Server-Side Request Forgery (SSRF)

LLaMA-Factory contains a server-side request forgery vulnerability in the OpenAI-compatible API multimodal media URL handler that allows unauthenticated attackers to bypass SSRF validation. The checkssrfurl guard validates URLs once but requests.get follows redirects and re-resolves DNS without...

PT-2026-85559

Name of the Vulnerable Software and Affected Versions OGX formerly Llama Stack versions up to commit fbe8e0f Description An unauthenticated server-side request forgery SSRF exists in the OpenAI-compatible 'POST /v1/responses' endpoint. The issue occurs because MCP tool definitions accept a server...

PT-2026-85566

LLaMA-Factory contains a server-side request forgery vulnerability in the OpenAI-compatible API multimodal media URL handler that allows unauthenticated attackers to bypass SSRF validation. The check ssrf url guard validates URLs once but requests.get follows redirects and re-resolves DNS without...

Trust Me, I'M Your Developer: Self-Issued Authentication in Large Language Models

Large language model LLM security has largely focused on role-playing jailbreaks, with less attention to what happens when a user asks an LLM to verify an identity claim through a test designed by the model itself. We study this behavior through a staged developer-identity experiment with ChatGPT...

Circuit Discovery Helps Detect LLM Jailbreaking: A Mechanistic Interpretability Study

Despite extensive safety alignment, large language models LLMs remain vulnerable to jailbreak attacks that bypass safeguards to elicit harmful content. While prior work attributes this vulnerability to safety training limitations, the internal mechanisms by which LLMs process adversarial prompts...

Prompt Structure Redistributes, Not Reduces: An Empirical Analysis of Security-Weaknesses in LLM-Generated Python Code

Large Language Models LLMs increasingly generate code from natural-language prompts, making prompt engineering a key mechanism for shaping the security of generated software. Structured and security-oriented prompts are widely used to encourage safer code, yet their effects extend beyond whether...

From Noise to Signal: Improving Security Log Anomaly Detection Using LLMs with Endpoint-Specific Logs

Existing approaches to anomalous behaviour log detection, such as Wazuh rely primarily on predefined detection rules, while statistical anomaly detection approaches such as OpenSearch identify deviations from previously observed behavioural patterns. Recent research has investigated LLMs for log...

CVE-2026-43631

A flaw was found in llama.cpp. This use-after-free vulnerability in the llama-server component, specifically within the vocab pointer when the --sleep-idle-seconds feature is enabled, allows unauthenticated remote attackers to execute arbitrary code. Attackers can exploit this by sending requests...

ag-llama-api (>=0.0.1, <=0.0.31), ag-llama-api-s (>=0.0.4, <=0.0.12) +23 more potentially affected by CVE-2026-69112 via accelerate (0.20.3)

accelerate PYPI version 0.20.3 is affected by a known vulnerability. The following packages have a transitive dependency on accelerate and may be impacted: - ag-llama-api =0.0.1, =0.0.4, =0.0.4, =0.0.11, =0.0.13, =0.0.15, =1.15.3, =1.16.3, =0.0.1, =2.0.0, =2.0.1.post2 - portfolio-agent 0.3.0rc1 a...

SUSE CVE-2026-43622

llama.cpp builds b1886 through b7445 contain a double free vulnerability in the LLaMA-Android JNI wrapper where new1batch allocates memory using malloc while free1batch deallocates it using the C++ delete operator, causing heap metadata corruption. Attackers can trigger this memory management...

SUSE CVE-2026-43628

llama.cpp builds b3978 through b9058 contain an integer underflow and out-of-bounds read vulnerability in the DRY sampler that allows unauthenticated attackers to trigger a heap buffer underflow by sending a crafted HTTP request with dryallowedlength set to INT32MIN to the /v1/completions or...

SUSE CVE-2026-43629

llama.cpp builds b4882 through b9058 contain a heap buffer overflow vulnerability in the KV cache state restore path where the statereaddata function computes write size without overflow checking, allowing attackers with write access to the slotsavepath directory to corrupt heap memory. Attackers...

SUSE CVE-2026-43630

llama.cpp builds b4882 before b8487 contain an out-of-bounds read vulnerability in the recurrent memory state restore path that allows attackers with write access to the slot save directory to read memory past the end of the allocated cells array. Attackers can craft a malicious slot file with an...

SUSE CVE-2026-43632

llama.cpp builds b7492 through the latest b9060 contains a use-after-free vulnerability in llama-server affecting six tokenization endpoints /tokenize, /detokenize, /infill, /apply-template, /rerank, and /anthropic/counttokens that bypass the task queue and access ctxserver.vocab directly on HTTP...

Exploit for CVE-2026-70638

CVE-2026-70638 - llama.cpp LLaMA-Android JNI Integer Overflow...

EUVD-2026-54276

llama.cpp builds b1886 through b7445 contain an integer overflow vulnerability in the LLaMA-Android JNI wrapper where the new1batch function multiplies sizeofllamaseqid by an attacker-controlled nseqmax parameter without overflow validation, causing heap buffer allocation to wrap and allocate...

EUVD-2026-54280

llama.cpp builds b1886 through b7445 contain a race condition use-after-free vulnerability in the LLaMA-Android JNI wrapper where bench1model and free1context lack synchronization, allowing Thread A to operate on freed memory while Thread B concurrently frees the llamacontext. Attackers can explo...

EUVD-2026-54286

llama.cpp builds b7492 through the latest b9060 contains a use-after-free vulnerability in the vocab pointer of llama-server when the --sleep-idle-seconds feature is enabled, allowing unauthenticated remote attackers to execute arbitrary code. Attackers can trigger the vulnerability by sending...

EUVD-2026-54287

llama.cpp builds b7492 through the latest b9060 contains a use-after-free vulnerability in llama-server affecting six tokenization endpoints /tokenize, /detokenize, /infill, /apply-template, /rerank, and /anthropic/counttokens that bypass the task queue and access ctxserver.vocab directly on HTTP...

Linux Distros Unpatched Vulnerability : CVE-2026-43628

The Linux/Unix host has one or more packages installed that are impacted by a vulnerability without a vendor supplied patch available. - llama.cpp builds b3978 through b9058 contain an integer underflow and out-of-bounds read vulnerability in the DRY sampler that allows unauthenticated attackers ...