LexiMark: Robust Watermarking via Lexical Substitutions to Enhance Membership Verification of an LLM's Textual Training Data

Large language models LLMs can be trained or fine-tuned on data obtained without the owner's consent. Verifying whether a specific LLM was trained on particular data instances or an entire dataset is extremely challenging. Dataset watermarking addresses this by embedding identifiable modification...

Unlearning-Enhanced Website Fingerprinting Attack: against Backdoor Poisoning in Anonymous Networks

Website Fingerprinting WF is an effective tool for regulating and governing the dark web. However, its performance can be significantly degraded by backdoor poisoning attacks in practical deployments. This paper aims to address the problem of hidden backdoor poisoning attacks faced by Website...

Certified Unlearning for Neural Networks

We address the problem of machine unlearning, where the goal is to remove the influence of specific training data from a model upon request, motivated by privacy concerns and regulatory requirements such as the "right to be forgotten." Unfortunately, existing methods rely on restrictive assumptio...

Private Memorization Editing: Turning Memorization into a Defense to Strengthen Data Privacy in Large Language Models

Large Language Models LLMs memorize, and thus, among huge amounts of uncontrolled data, may memorize Personally Identifiable Information PII, which should not be stored and, consequently, not leaked. In this paper, we introduce Private Memorization Editing PME, an approach for preventing private...

What Really Is a Member? Discrediting Membership Inference Via Poisoning

Membership inference tests aim to determine whether a particular data point was included in a language model's training set. However, recent works have shown that such tests often fail under the strict definition of membership based on exact matching, and have suggested relaxing this definition t...

VLMs Can Aggregate Scattered Training Patches

Whitepaper called VLMs Can Aggregate Scattered Training Patches...

Towards Secure MLOps: Surveying Attacks, Mitigation Strategies, and Research Challenges

The rapid adoption of machine learning ML technologies has driven organizations across diverse sectors to seek efficient and reliable methods to accelerate model development-to-deployment. Machine Learning Operations MLOps has emerged as an integrative approach addressing these requirements by...

Private LoRA Fine-Tuning of Open-Source LLMs with Homomorphic Encryption

Preserving data confidentiality during the fine-tuning of open-source Large Language Models LLMs is crucial for sensitive applications. This work introduces an interactive protocol adapting the Low-Rank Adaptation LoRA technique for private fine-tuning. Homomorphic Encryption HE protects the...

BadMoE: Backdooring Mixture-Of-Experts LLMs Via Optimizing Routing Triggers and Infecting Dormant Experts

Mixture-of-Experts MoE have emerged as a powerful architecture for large language models LLMs, enabling efficient scaling of model capacity while maintaining manageable computational costs. The key advantage lies in their ability to route different tokens to different "expert'' networks within th...

Leveraging LLM to Strengthen ML-Based Cross-Site Scripting Detection

According to the Open Web Application Security Project OWASP, Cross-Site Scripting XSS is a critical security vulnerability. Despite decades of research, XSS remains among the top 10 security vulnerabilities. Researchers have proposed various techniques to protect systems from XSS attacks, with...

AI Industry is Trying to Subvert the Definition of “Open Source AI”

The Open Source Initiative has published news article here its definition of "open source AI," and it's terrible. It allows for secret training data and mechanisms. It allows for development to be done in secret. Since for a neural network, the training data is the source code--it's how the model...

Reducing False Positives in API Security: Advanced Techniques Using Machine Learning

False positives in API security are a serious problem, often resulting in wasted results and time, missing real threats, alert fatigue, and operational disruption. Fortunately, however, emerging technologies like machine learning ML can help organizations minimize false positives and streamline t...

AI Pulse: Brazil Gets Bold with Meta, Interpol’s Red Flag & more

The second edition of AI Pulse is all about AI regulation: what’s coming, why it matters, and what might happen without it. We look at Brazil’s hard não to Meta, how communities are pushing back against AI training data use, Interpol’s warnings about AI deepfakes, and more...

Peloton accused of providing customer chat data to train AI

It seems that Peloton may have been providing more training than just for its customers, as its set to face court in California accused of using user chat data to train AI. Peloton Interactive, Inc. is a US-based exercise equipment and media company, known for its stationary bicycles, treadmills,...

SUSE CVE-2024-5206

A sensitive data leakage vulnerability was identified in scikit-learn's TfidfVectorizer, specifically in versions up to and including 1.4.1.post1, which was fixed in version 1.5.0. The vulnerability arises from the unexpected storage of all tokens present in the training data within the stopwords...

CVE-2024-5206

A sensitive data leakage vulnerability was identified in scikit-learn's TfidfVectorizer, specifically in versions up to and including 1.4.1.post1, which was fixed in version 1.5.0. The vulnerability arises from the unexpected storage of all tokens present in the training data within the stopwords...

CVE-2024-5206 Sensitive Data Leakage in sklearn.feature_extraction.text.TfidfVectorizer in scikit-learn/scikit-learn

A sensitive data leakage vulnerability was identified in scikit-learn's TfidfVectorizer, specifically in versions up to and including 1.4.1.post1, which was fixed in version 1.5.0. The vulnerability arises from the unexpected storage of all tokens present in the training data within the stopwords...

CVE-2024-5206 Sensitive Data Leakage in sklearn.feature_extraction.text.TfidfVectorizer in scikit-learn/scikit-learn

A sensitive data leakage vulnerability was identified in scikit-learn's TfidfVectorizer, specifically in versions up to and including 1.4.1.post1, which was fixed in version 1.5.0. The vulnerability arises from the unexpected storage of all tokens present in the training data within the stopwords...

scikit-learn Security Vulnerabilities

scikit-learn is an open source Python-based machine learning package that supports spam detection, image recognition, and prediction of continuous-valued attributes of associations. A security vulnerability exists in scikit-learn 1.4.1.post1 and earlier versions, which stems from accidentally...

Securing AI Development in the Cloud: Navigating the Risks and Opportunities

AI-TRiSM - Trust, Risk and Security Management in the Age of AI Co-authored by Lara Sunday and Pojan Shahrivar As artificial intelligence AI and machine learning ML technologies continue to advance and proliferate, organizations across industries are investing heavily in these transformative...