Tags → #data poisoning

Data Poisoning Attacks Impact Local Differential Privacy Protocols

The increasing threat of data poisoning attacks in cybersecurity highlights the need for effective countermeasures and adaptive detection mechanisms to protect machine learning systems and sensitive data.
Vulnerabilities in Large Language Models: Data Poisoning Threats

Recent research highlights vulnerabilities in large language models, particularly concerning data poisoning and jailbreak-tuning, emphasizing the need for improved data integrity and security measures.
Large Language Models Used in Hardware Design Raise Security Concerns

Large language models are being increasingly used in hardware design for generating hardware description language code, but their integration raises significant security concerns related to data poisoning and backdoor attacks.
Study Reveals New Method for Backdoor Attacks on DNNs

A recent study has introduced Clean-Label Physical Backdoor Attacks (CLPBA), a method that employs physical triggers to manipulate machine learning models while maintaining the original labels of poisoned samples.
Backdoor Attacks on Deep Learning Systems: New Toolkit Introduced

The article discusses the increasing significance of backdoor attacks on deep learning systems and introduces BackdoorMBTI, a toolkit designed to evaluate these attacks across multiple data modalities.