Safe reinforcement learning (Safe RL) aims to ensure policy performance while satisfying safety constraints. However, most existing Safe RL methods assume benign environments, making them vulnerable to adversarial perturbations commonly encountered in real-world settings. In addition, existing gradient-based adversarial attacks typically require access to the policy's gradient information, which is often impractical in real-world scenarios. To address these challenges, we propose an adversarial attack framework to reveal vulnerabilities of Safe RL policies. Using expert demonstrations and black-box environment interaction, our framework learns a constraint model and a surrogate (learner) policy, enabling gradient-based attack optimization without requiring the victim policy's internal gradients or the ground-truth safety constraints. We further provide theoretical analysis establishing feasibility and deriving perturbation bounds. Experiments on multiple Safe RL benchmarks demonstrate the effectiveness of our approach under limited privileged access.

Federated learning (FL) enables collaborative training across organizational silos without sharing raw data, making it attractive for privacy-sensitive applications. With the rapid adoption of large language models (LLMs), federated fine-tuning of generative LLMs has gained attention as a way to leverage distributed data while preserving confidentiality. However, this setting introduces fundamental challenges: (i) privacy leakage of personally identifiable information (PII) due to LLM memorization, and (ii) a persistent tension between global generalization and local utility under heterogeneous data. Existing defenses, such as data sanitization and differential privacy, reduce leakage but often degrade downstream performance. We propose SecureGate, a privacy-aware federated fine-tuning framework for LLMs that provides fine-grained privacy control without sacrificing utility. SecureGate employs a dual-adapter LoRA architecture: a secure adapter that learns sanitized, globally shareable representations, and a revealing adapter that captures sensitive, organization-specific knowledge. A token-controlled gating module selectively activates these adapters at inference time, enabling controlled information disclosure without retraining. Extensive experiments across multiple LLMs and real-world datasets show that SecureGate improves task utility while substantially reducing PII leakage, achieving up to a 31.66X reduction in inference attack accuracy and a 17.07X reduction in extraction recall for unauthorized requests. Additionally, it maintains 100% routing reliability to the correct adapter and incurs only minimal computational and communication overhead.

Membership inference attack (MIA) poses a significant privacy threat in federated learning (FL) as it allows adversaries to determine whether a client's private dataset contains a specific data sample. While defenses against membership inference attacks in standard FL have been well studied, the recent shift toward federated fine-tuning has introduced new, largely unexplored attack surfaces. To highlight this vulnerability in the emerging FL paradigm, we demonstrate that federated prompt-tuning, which adapts pre-trained models with small input prefixes to improve efficiency, also exposes a new vector for privacy attacks. We propose PromptMIA, a membership inference attack tailored to federated prompt-tuning, in which a malicious server can insert adversarially crafted prompts and monitors their updates during collaborative training to accurately determine whether a target data point is in a client's private dataset. We formalize this threat as a security game and empirically show that PromptMIA consistently attains high advantage in this game across diverse benchmark datasets. Our theoretical analysis further establishes a lower bound on the attack's advantage which explains and supports the consistently high advantage observed in our empirical results. We also investigate the effectiveness of standard membership inference defenses originally developed for gradient or output based attacks and analyze their interaction with the distinct threat landscape posed by PromptMIA. The results highlight non-trivial challenges for current defenses and offer insights into their limitations, underscoring the need for defense strategies that are specifically tailored to prompt-tuning in federated settings.

Multimodal generative AI systems like Stable Diffusion, DALL-E, and MidJourney have fundamentally changed how synthetic images are created. These tools drive innovation but also enable the spread of misleading content, false information, and manipulated media. As generated images become harder to distinguish from photographs, detecting them has become an urgent priority. To combat this challenge, We release MS COCOAI, a novel dataset for AI generated image detection consisting of 96000 real and synthetic datapoints, built using the MS COCO dataset. To generate synthetic images, we use five generators: Stable Diffusion 3, Stable Diffusion 2.1, SDXL, DALL-E 3, and MidJourney v6. Based on the dataset, we propose two tasks: (1) classifying images as real or generated, and (2) identifying which model produced a given synthetic image. The dataset is available at https://huggingface.co/datasets/Rajarshi-Roy-research/Defactify_Image_Dataset.

The rise of space AI is reshaping government and industry through applications such as disaster detection, border surveillance, and climate monitoring, powered by massive data from commercial and governmental low Earth orbit (LEO) satellites. Federated satellite learning (FSL) enables joint model training without sharing raw data, but suffers from slow convergence due to intermittent connectivity and introduces critical trust challenges--where biased or falsified updates can arise across satellite constellations, including those injected through cyberattacks on inter-satellite or satellite-ground communication links. We propose OrbitChain, a blockchain-backed framework that empowers trustworthy multi-vendor collaboration in LEO networks. OrbitChain (i) offloads consensus to high-altitude platforms (HAPs) with greater computational capacity, (ii) ensures transparent, auditable provenance of model updates from different orbits owned by different vendors, and (iii) prevents manipulated or incomplete contributions from affecting global FSL model aggregation. Extensive simulations show that OrbitChain reduces computational and communication overhead while improving privacy, security, and global model accuracy. Its permissioned proof-of-authority ledger finalizes over 1000 blocks with sub-second latency (0.16,s, 0.26,s, 0.35,s for 1-of-5, 3-of-5, and 5-of-5 quorums). Moreover, OrbitChain reduces convergence time by up to 30 hours on real satellite datasets compared to single-vendor, demonstrating its effectiveness for real-time, multi-vendor learning. Our code is available at https://github.com/wsu-cyber-security-lab-ai/OrbitChain.git

Underground mining operations depend on sensor networks to monitor critical parameters such as temperature, gas concentration, and miner movement, enabling timely hazard detection and safety decisions. However, transmitting raw sensor data to a centralized server for machine learning (ML) model training raises serious privacy and security concerns. Federated Learning (FL) offers a promising alternative by enabling decentralized model training without exposing sensitive local data. Yet, applying FL in underground mining presents unique challenges: (i) Adversaries may eavesdrop on shared model updates to launch model inversion or membership inference attacks, compromising data privacy and operational safety; (ii) Non-IID data distributions across mines and sensor noise can hinder model convergence. To address these issues, we propose FedMining--a privacy-preserving FL framework tailored for underground mining. FedMining introduces two core innovations: (1) a Decentralized Functional Encryption (DFE) scheme that keeps local models encrypted, thwarting unauthorized access and inference attacks; and (2) a balancing aggregation mechanism to mitigate data heterogeneity and enhance convergence. Evaluations on real-world mining datasets demonstrate FedMining's ability to safeguard privacy while maintaining high model accuracy and achieving rapid convergence with reduced communication and computation overhead. These advantages make FedMining both secure and practical for real-time underground safety monitoring.

Secure aggregation is a common technique in federated learning (FL) for protecting data privacy from both curious internal entities (clients or server) and external adversaries (eavesdroppers). However, in dynamic and resource-constrained environments such as low Earth orbit (LEO) satellite networks, traditional secure aggregation methods fall short in two aspects: (1) they assume continuous client availability while LEO satellite visibility is intermittent and irregular; (2) they consider privacy in each communication round but have overlooked the possible privacy leakage through multiple rounds. To address these limitations, we propose LTP-FLEO, an asynchronous FL framework that preserves long-term privacy (LTP) for LEO satellite networks. LTP-FLEO introduces (i) privacy-aware satellite partitioning, which groups satellites based on their predictable visibility to the server and enforces joint participation; (ii) model age balancing, which mitigates the adverse impact of stale model updates; and (iii) fair global aggregation, which treats satellites of different visibility durations in an equitable manner. Theoretical analysis and empirical validation demonstrate that LTP-FLEO effectively safeguards both model and data privacy across multi-round training, promotes fairness in line with satellite contributions, accelerates global convergence, and achieves competitive model accuracy.

Underground mining operations rely on distributed sensor networks to collect critical data daily, including mine temperature, toxic gas concentrations, and miner movements for hazard detection and operational decision-making. However, transmitting raw sensor data to a central server for training deep learning models introduces significant privacy risks, potentially exposing sensitive mine-specific information. Federated Learning (FL) offers a transformative solution by enabling collaborative model training while ensuring that raw data remains localized at each mine. Despite its advantages, FL in underground mining faces key challenges: (i) An attacker may compromise a mine's local model by employing techniques such as sign-flipping attacks or additive noise, leading to erroneous predictions; (ii) Low-quality (yet potentially valuable) data, caused by poor lighting conditions or sensor inaccuracies in mines may degrade the FL training process. In response, this paper proposes MineDetect, a defense FL framework that detects and isolates the attacked models while mitigating the impact of mines with low-quality data. MineDetect introduces two key innovations: (i) Detecting attacked models (maliciously manipulated) by developing a history-aware mechanism that leverages local and global averages of gradient updates; (ii) Identifying and eliminating adversarial influences from unreliable models (generated by clients with poor data quality) on the FL training process. Comprehensive simulations across diverse datasets demonstrate that MineDetect outperforms existing methods in both robustness and accuracy, even in challenging non-IID data scenarios. Its ability to counter adversarial influences while maintaining lower computational efficiency makes it a vital advancement for improving safety and operational effectiveness in underground mining.