Unveiling Privacy Risks in Multi-Modal Large Language Models: Task-Specific Vulnerabilities and Mitigation Challenges

Privacy risks in text-only Large Language Models LLMs are well studied, particularly their tendency to memorize and leak sensitive information. However, Multi-modal Large Language Models MLLMs, which process both text and images, introduce unique privacy challenges that remain underexplored...

EUVD-2003-0007

Malware in sbrugna...

EUVD-2023-38044

Malicious code in bioql PyPI...

Code Injection Vulnerabilities in Various ABB Products

ABB ASPECT-Enterprise is a scalable building energy management and control solution.ABB NEXUS Series is a monitoring and control management system.ABB MATRIX Series is an embedded IoT ASPECT control engine designed to provide flexible field control for medium to large field control applications...

CVE-2020-8576

Clustered Data ONTAP versions prior to 9.3P19, 9.5P14, 9.6P9 and 9.7 are susceptible to a vulnerability which when successfully exploited could lead to addition or modification of data or disclosure of sensitive information...

CVE-2020-1806

Huawei Honor V10 smartphones with versions earlier than 10.0.0.156C00E156R2P4 has three out of bounds vulnerabilities. Certain driver program does not sufficiently validate certain parameters received, that would lead to several bytes out of bound read. Successful exploit may cause information...

PT-2025-20899 · Amd · Amd Uprof

Name of the Vulnerable Software and Affected Versions: AMD uProf affected versions not specified Description: A junction point vulnerability within AMD uProf can allow a local low-privileged attacker to create junction points, potentially resulting in arbitrary file deletion or disclosure...

CVE-2024-20680 Windows Message Queuing Client (MSMQC) Information Disclosure

...