The Trigger in the Haystack: Extracting and Reconstructing LLM Backdoor Triggers

Detecting whether a model has been poisoned is a longstanding problem in AI security. In this work, we present a practical scanner for identifying sleeper agent-style backdoors in causal language models. Our approach relies on two key findings: first, sleeper agents tend to memorize poisoning dat...

PT-2026-5654

Name of the Vulnerable Software and Affected Versions huggingface/text-generation-inference version 3.3.6 huggingface/text-generation-inference versions prior to 3.3.7 Description A flaw exists in huggingface/text-generation-inference that allows unauthenticated remote attackers to cause a...

Text Generation Inference 资源管理错误漏洞

Text Generation Inference is a Rust, Python, and gRPC server developed by Hugging Face for text generation inference. Version 3.3.6 of Text Generation Inference contains a resource management vulnerability. This vulnerability stems from the unlimited acquisition of external images during input...

Evaluating Large Language Models for Security Bug Report Prediction

Early detection of security bug reports SBRs is critical for timely vulnerability mitigation. We present an evaluation of prompt-based engineering and fine-tuning approaches for predicting SBRs using Large Language Models LLMs. Our findings reveal a distinct trade-off between the two approaches...

Timing Attack

OctoPrint is vulnerable to Timing Attack. The vulnerability is due to character-by-character API key comparison with early termination, which allows a network-based attacker to infer valid API keys by measuring response times and guessing the key one character at a time...

Hardware-Triggered Backdoors

Machine learning models are routinely deployed on a wide range of computing hardware. Although such hardware is typically expected to produce identical results, differences in its design can lead to small numerical variations during inference. In this work, we show that these variations can be...

SQLMAP-CTT-v2.0-33-Layer-Fractal-Resonance-SQL-Injection-Engine

SQLMAP-CTT-v2.0-33-Layer-Fractal-Resonance-SQL...

CVE-2025-22234

The fix applied in CVE-2025-22228 inadvertently broke the timing attack mitigation implemented in DaoAuthenticationProvider. This can allow attackers to infer valid usernames or other authentication behavior via response-time differences under certain configurations. Mitigation Mitigation for thi...

Spring Security has a broken timing attack mitigation implemented in DaoAuthenticationProvide

The fix applied in CVE-2025-22228 inadvertently broke the timing attack mitigation implemented in DaoAuthenticationProvider. This can allow attackers to infer valid usernames or other authentication behavior via response-time differences under certain configurations...

CVE-2025-22234

The fix applied in CVE-2025-22228 inadvertently broke the timing attack mitigation implemented in DaoAuthenticationProvider. This can allow attackers to infer valid usernames or other authentication behavior via response-time differences under certain configurations...

CVE-2025-22234

CVE-2025-22234 is associated with Spring Security’s timing-attack mitigation in DaoAuthenticationProvider. The described issue states that the fix applied in CVE-2025-22228 accidentally broke the mitigation, enabling an attacker to infer usernames or authentication behavior via response-time diff...

EUVD-2026-3787

The fix applied in CVE-2025-22228 inadvertently broke the timing attack mitigation implemented in DaoAuthenticationProvider. This can allow attackers to infer valid usernames or other authentication behavior via response-time differences under certain configurations...

PT-2026-3934

Neo4j Enterprise edition versions prior to 2025.11.2 and 5.26.17 are vulnerable to a potential information disclosure by an attacker who has some legitimate access to the database. The vulnerability allows attacker without read access to a property to infer information about its value by trying t...

CVE-2026-23996

CVE-2026-23996 concerns the FastAPI Api Key library. Version 1.1.0 is reported to expose a timing side-channel in verify_key(), where a random delay is applied only on verification failures. This enables an attacker to statistically distinguish valid from invalid API keys by measuring response la...

Unity Linux 20.1060e / 20.1070e Security Update: kernel (UTSA-2026-004086)

The Unity Linux 20 host has a package installed that is affected by a vulnerability as referenced in the UTSA-2026-004086 advisory. The flowdissector feature in the Linux kernel 4.3 through 5.x before 5.3.10 has a device tracking vulnerability, aka CID-55667441c84f. This occurs because the auto...

LoRA As Oracle

Backdoored and privacy-leaking deep neural networks pose a serious threat to the deployment of machine learning systems in security-critical settings. Existing defenses for backdoor detection and membership inference typically require access to clean reference models, extensive retraining, or...

Unity Linux 20.1050e / 20.1060e / 20.1070e Security Update: kernel (UTSA-2026-001603)

The Unity Linux 20 host has a package installed that is affected by a vulnerability as referenced in the UTSA-2026-001603 advisory. System software utilizing Lazy FP state restore technique on systems using Intel Core-based microprocessors may potentially allow a local process to infer data from...

acido (=0.15.0), adstoolbox (>=2025.12.2.2 <=2026.5.19) +205 more potentially affected by CVE-2026-21226 via azure-core (>=1.10.0 <=1.37.0)

azure-core PYPI version =1.10.0, =2025.12.2.2, =0.1.12, =0.1.31, =0.1.1, =0.0.2, =0.0.53, =0.1.0, =0.9.0, =0.2.100, =0.2.123, =1.0.0, =1.0.0, =0.1.0b1, =0.1.0b2 and more Source cves: CVE-2026-21226 Source advisory: OSV:GHSA-JM66-CG57-JJV5...

CVE-2021-33649

When performing the inference shape operation of the Transpose operator, if the value in the perm element is greater than or equal to the size of the inputshape, it will access data outside of bounds of inputshape which allocated from heap buffers...

CVE-2021-41216

TensorFlow is an open source platform for machine learning. In affected versions the shape inference function for Transpose is vulnerable to a heap buffer overflow. This occurs whenever perm contains negative elements. The shape inference function does not validate that the indices in perm are al...