GuardReasoner-VL: Safeguarding VLMs Via Reinforced Reasoning

To enhance the safety of VLMs, this paper introduces a novel reasoning-based VLM guard model dubbed GuardReasoner-VL. The core idea is to incentivize the guard model to deliberatively reason before making moderation decisions via online RL. First, we construct GuardReasoner-VLTrain, a reasoning...

Unveiling the Black Box: a Multi-Layer Framework for Explaining Reinforcement Learning-Based Cyber Agents

Reinforcement Learning RL agents are increasingly used to simulate sophisticated cyberattacks, but their decision-making processes remain opaque, hindering trust, debugging, and defensive preparedness. In high-stakes cybersecurity contexts, explainability is essential for understanding how...

Adaptive Security Policy Management in Cloud Environments Using Reinforcement Learning

The security of cloud environments, such as Amazon Web Services AWS, is complex and dynamic. Static security policies have become inadequate as threats evolve and cloud resources exhibit elasticity 1. This paper addresses the limitations of static policies by proposing a security policy managemen...

Improved Algorithms for Differentially Private Language Model Alignment

Language model alignment is crucial for ensuring that large language models LLMs align with human preferences, yet it often involves sensitive user data, raising significant privacy concerns. While prior work has integrated differential privacy DP with alignment techniques, their performance...

Remote Rowhammer Attack Using Adversarial Observations on Federated Learning Clients

Federated Learning FL has the potential for simultaneous global learning amongst a large number of parallel agents, enabling emerging AI such as LLMs to be trained across demographically diverse data. Central to this being efficient is the ability for FL to perform sparse gradient updates and...

Large Language Models Are Autonomous Cyber Defenders

Fast and effective incident response is essential to prevent adversarial cyberattacks. Autonomous Cyber Defense ACD aims to automate incident response through Artificial Intelligence AI agents that plan and execute actions. Most ACD approaches focus on single-agent scenarios and leverage...

DMRL: Data- and Model-Aware Reward Learning for Data Extraction

Large language models LLMs are inherently vulnerable to unintended privacy breaches. Consequently, systematic red-teaming research is essential for developing robust defense mechanisms. However, current data extraction methods suffer from several limitations: 1 rely on dataset duplicates...

The Steganographic Potentials of Language Models

The potential for large language models LLMs to hide messages within plain text steganography poses a challenge to detection and thwarting of unaligned AI agents, and undermines faithfulness of LLMs reasoning. We explore the steganographic capabilities of LLMs fine-tuned via reinforcement learnin...

Modeling Behavioral Preferences of Cyber Adversaries Using Inverse Reinforcement Learning

This paper presents a holistic approach to attacker preference modeling from system-level audit logs using inverse reinforcement learning IRL. Adversary modeling is an important capability in cybersecurity that lets defenders characterize behaviors of potential attackers, which enables attributio...

Application of Deep Reinforcement Learning for Intrusion Detection in Internet of Things: a Systematic Review

The Internet of Things IoT has significantly expanded the digital landscape, interconnecting an unprecedented array of devices, from home appliances to industrial equipment. This growth enhances functionality, e.g., automation, remote monitoring, and control, and introduces substantial security...

Evaluating the Effectiveness of Reward Modeling of Generative AI Systems

New research evaluating the effectiveness of reward modeling during Reinforcement Learning from Human Feedback RLHF: "SEAL: Systematic Error Analysis for Value ALignment." The paper introduces quantitative metrics for evaluating the effectiveness of modeling and aligning human values: Abstract:...

Poisoning AI Models

New research into poisoning AI models: The researchers first trained the AI models using supervised learning and then used additional "safety training" methods, including more supervised learning, reinforcement learning, and adversarial training. After this, they checked if the AI still had hidde...

AutoPentest-DRL - Automated Penetration Testing Using Deep Reinforcement Learning

AutoPentest-DRL is an automated penetration testing framework based on Deep Reinforcement Learning DRL techniques. The framework determines the most appropriate attack path for a given network, and can be used to execute a simulated attack on that network via penetration testing tools, such as...

CyberBattleSim - An Experimentation And Research Platform To Investigate The Interaction Of Automated Agents In An Abstract Simulated Network Environments

CyberBattleSim is an experimentation research platform to investigate the interaction of automated agents operating in a simulated abstract enterprise network environment. The simulation provides a high-level abstraction of computer networks and cyber security concepts. Its Python-based Open AI G...

Gamifying machine learning for stronger security and AI models

To stay ahead of adversaries, who show no restraint in adopting tools and techniques that can help them attain their goals, Microsoft continues to harness AI and machine learning to solve security challenges. One area we’ve been experimenting on is autonomous systems. In a simulated enterprise...

Gamifying machine learning for stronger security and AI models

To stay ahead of adversaries, who show no restraint in adopting tools and techniques that can help them attain their goals, Microsoft continues to harness AI and machine learning to solve security challenges. One area we’ve been experimenting on is autonomous systems. In a simulated enterprise...

Pesidious - Malware Mutation Using Reinforcement Learning And Generative Adversarial Networks

Malware Mutation using Deep Reinforcement Learning and GANs The purpose of the tool is to use artificial intelligence to mutate a malware PE32 only sample to bypass AI powered classifiers while keeping its functionality intact. In the past, notable work has been done in this domain with researche...

Manipulating Machine Learning Systems by Manipulating Training Data

Interesting research: "TrojDRL: Trojan Attacks on Deep Reinforcement Learning Agents": Abstract:: Recent work has identified that classification models implemented as neural networks are vulnerable to data-poisoning and Trojan attacks at training time. In this work, we show that these training-ti...

Automatic Machine Learning Penetration Test Tool: Deep Exploit

DeepExploit is fully automated penetration tool linked with Metasploit. It identifies the status of all opened ports on the target server and executes the exploit at pinpoint using Machine Learning. DeepExploit consists of the machine learning model A3C and Metasploit . The A3C executes exploit t...

Application Security Testing — The Wallarm Approach

Testing the security of the corporate applications is a part of every-day life for Ops and DevOps professionals. Larger companies have whole teams dedicated to independent security testing, called Red Teams. These folks use various tools at their disposal to discover the flaws in both application...