Papers

Power grids increasingly need real-time situational awareness under the ever-evolving cyberthreat landscape. Advances in snapshot-based system identification approaches have enabled accurately estimating states and topology from a snapshot of measurement data, under random bad data and topology errors. However, modern interactive, targeted false data can stay undetectable to these methods, and significantly compromise estimation accuracy. This work advances system identification that combines snapshot-based method with time-series model via Bayesian Integration, to advance cyber resiliency against both random and targeted false data. Using a distance-based time-series model, this work can leverage historical data of different distributions induced by changes in grid topology and other settings. The normal system behavior captured from historical data is integrated into system identification through a Bayesian treatment, to make solutions robust to targeted false data. We experiment on mixed random anomalies (bad data, topology error) and targeted false data injection attack (FDIA) to demonstrate our method's 1) cyber resilience: achieving over 70% reduction in estimation error under FDIA; 2) anomalous data identification: being able to alarm and locate anomalous data; 3) almost linear scalability: achieving comparable speed with the snapshot-based baseline, both taking <1min per time tick on the large 2,383-bus system using a laptop CPU.

LLM-integrated applications are vulnerable to prompt injection attacks, where an attacker contaminates the input to inject malicious instructions, causing the LLM to follow the attacker's intent instead of the original user's. Existing prompt injection detection methods often have sub-optimal performance and/or high computational overhead. In this work, we propose PIShield, an effective and efficient detection method based on the observation that instruction-tuned LLMs internally encode distinguishable signals for prompts containing injected instructions. PIShield leverages residual-stream representations and a simple linear classifier to detect prompt injection, without expensive model fine-tuning or response generation. We conduct extensive evaluations on a diverse set of short- and long-context benchmarks. The results show that PIShield consistently achieves low false positive and false negative rates, significantly outperforming existing baselines. These findings demonstrate that internal representations of instruction-tuned LLMs provide a powerful and practical foundation for prompt injection detection in real-world applications.

Internet of Things (IoT) networks generate diverse and high-volume traffic that reflects both normal activity and potential threats. Deriving meaningful insight from such telemetry requires cross-layer interpretation of behaviors, protocols, and context rather than isolated detection. This work presents an LLM-powered AI agent framework that converts raw packet captures into structured and semantically enriched representations for interactive analysis. The framework integrates feature extraction, transformer-based anomaly detection, packet and flow summarization, threat intelligence enrichment, and retrieval-augmented question answering. An AI agent guided by a large language model performs reasoning over the indexed traffic artifacts, assembling evidence to produce accurate and human-readable interpretations. Experimental evaluation on multiple IoT captures and six open models shows that hybrid retrieval, which combines lexical and semantic search with reranking, substantially improves BLEU, ROUGE, METEOR, and BERTScore results compared with dense-only retrieval. System profiling further indicates low CPU, GPU, and memory overhead, demonstrating that the framework achieves holistic and efficient interpretation of IoT network traffic.

With the rapid adoption of diffusion models for visual content generation, proving authorship and protecting copyright have become critical. This challenge is particularly important when model owners keep their models private and may be unwilling or unable to handle authorship issues, making third-party verification essential. A natural solution is to embed watermarks for later verification. However, existing methods require access to model weights and rely on computationally heavy procedures, rendering them impractical and non-scalable. To address these challenges, we propose NoisePrints, a lightweight watermarking scheme that utilizes the random seed used to initialize the diffusion process as a proof of authorship without modifying the generation process. Our key observation is that the initial noise derived from a seed is highly correlated with the generated visual content. By incorporating a hash function into the noise sampling process, we further ensure that recovering a valid seed from the content is infeasible. We also show that sampling an alternative seed that passes verification is infeasible, and demonstrate the robustness of our method under various manipulations. Finally, we show how to use cryptographic zero-knowledge proofs to prove ownership without revealing the seed. By keeping the seed secret, we increase the difficulty of watermark removal. In our experiments, we validate NoisePrints on multiple state-of-the-art diffusion models for images and videos, demonstrating efficient verification using only the seed and output, without requiring access to model weights.

This paper introduces a methodology based on Euclidean information theory to investigate local properties of secure communication over discrete memoryless wiretap channels. We formulate a constrained optimization problem that maximizes a legitimate user's information rate while imposing explicit upper bounds on both the information leakage to an eavesdropper and the informational cost of encoding the secret message. By leveraging local geometric approximations, this inherently non-convex problem is transformed into a tractable quadratic programming structure. It is demonstrated that the optimal Lagrange multipliers governing this approximated problem can be found by solving a linear program. The constraints of this linear program are derived from Karush-Kuhn-Tucker conditions and are expressed in terms of the generalized eigenvalues of channel-derived matrices. This framework facilitates the derivation of an analytical formula for an approximate local secrecy capacity. Furthermore, we define and analyze a new class of secret local contraction coefficients. These coefficients, characterized as the largest generalized eigenvalues of a matrix pencil, quantify the maximum achievable ratio of approximate utility to approximate leakage, thus measuring the intrinsic local leakage efficiency of the channel. We establish bounds connecting these local coefficients to their global counterparts defined over true mutual information measures. The efficacy of the proposed framework is demonstrated through detailed analysis and numerical illustrations for both general multi-mode channels and the canonical binary symmetric wiretap channel.

Large Language Model (LLM) based agents integrated into web browsers (often called agentic AI browsers) offer powerful automation of web tasks. However, they are vulnerable to indirect prompt injection attacks, where malicious instructions hidden in a webpage deceive the agent into unwanted actions. These attacks can bypass traditional web security boundaries, as the AI agent operates with the user privileges across sites. In this paper, we present a novel fuzzing framework that runs entirely in the browser and is guided by an LLM to automatically discover such prompt injection vulnerabilities in real time.

Managing passwords securely and conveniently is still an open problem for many users. Existing research has examined users' password management strategies and identified pain points, such as security concerns, leading to insecure practices. We investigate how Blind and Low-Vision (BLV) users tackle this problem and how password managers can assist them. This paper presents the results of a qualitative interview study with N = 33 BLV participants. We found that all participants utilize password managers to some extent, which they perceive as fairly accessible. However, the adoption is mainly driven by the convenience of storing and retrieving passwords. The security advantages - generating strong, random passwords - were avoided mainly due to the absence of practical accessibility. Password managers do not adhere to BLV users' underlying needs for agency, which stem from experiences with inaccessible software and vendors who deprioritize accessibility issues. Underutilization of password managers leads BLV users to adopt insecure practices, such as reusing predictable passwords or resorting to 'security through obscurity' by writing important credentials in braille. We conclude our analysis by discussing the need to implement practical accessibility and usability improvements for password managers as a way of establishing trust and secure practices while maintaining BLV users' agency.

Large language models (LLMs) with Mixture-of-Experts (MoE) architectures achieve impressive performance and efficiency by dynamically routing inputs to specialized subnetworks, known as experts. However, this sparse routing mechanism inherently exhibits task preferences due to expert specialization, introducing a new and underexplored vulnerability to backdoor attacks. In this work, we investigate the feasibility and effectiveness of injecting backdoors into MoE-based LLMs by exploiting their inherent expert routing preferences. We thus propose BadSwitch, a novel backdoor framework that integrates task-coupled dynamic trigger optimization with a sensitivity-guided Top-S expert tracing mechanism. Our approach jointly optimizes trigger embeddings during pretraining while identifying S most sensitive experts, subsequently constraining the Top-K gating mechanism to these targeted experts. Unlike traditional backdoor attacks that rely on superficial data poisoning or model editing, BadSwitch primarily embeds malicious triggers into expert routing paths with strong task affinity, enabling precise and stealthy model manipulation. Through comprehensive evaluations across three prominent MoE architectures (Switch Transformer, QwenMoE, and DeepSeekMoE), we demonstrate that BadSwitch can efficiently hijack pre-trained models with up to 100% success rate (ASR) while maintaining the highest clean accuracy (ACC) among all baselines. Furthermore, BadSwitch exhibits strong resilience against both text-level and model-level defense mechanisms, achieving 94.07% ASR and 87.18% ACC on the AGNews dataset. Our analysis of expert activation patterns reveals fundamental insights into MoE vulnerabilities. We anticipate this work will expose security risks in MoE systems and contribute to advancing AI safety.

Machine learning models are often vulnerable to inference attacks that expose sensitive information from their training data. Shadow model technique is commonly employed in such attacks, such as membership inference. However, the need for a large number of shadow models leads to high computational costs, limiting their practical applicability. Such inefficiency mainly stems from the independent training and use of these shadow models. To address this issue, we present a novel shadow pool training framework SHAPOOL, which constructs multiple shared models and trains them jointly within a single process. In particular, we leverage the Mixture-of-Experts mechanism as the shadow pool to interconnect individual models, enabling them to share some sub-networks and thereby improving efficiency. To ensure the shared models closely resemble independent models and serve as effective substitutes, we introduce three novel modules: path-choice routing, pathway regularization, and pathway alignment. These modules guarantee random data allocation for pathway learning, promote diversity among shared models, and maintain consistency with target models. We evaluate SHAPOOL in the context of various membership inference attacks and show that it significantly reduces the computational cost of shadow model construction while maintaining comparable attack performance.

Service Level Agreement (SLA) monitoring in service-oriented environments suffers from inherent trust conflicts when providers self-report metrics, creating incentives to underreport violations. We introduce a framework for generating verifiable SLA violation claims through trusted hardware monitors and zero-knowledge proofs, establishing cryptographic foundations for genuine trustworthiness in service ecosystems. Our approach starts with machine-readable SLA clauses converted into verifiable predicates and monitored within Trusted Execution Environments. These monitors collect timestamped telemetry, organize measurements into Merkle trees, and produce signed attestations. Zero-knowledge proofs aggregate Service-Level Indicators to evaluate compliance, generating cryptographic proofs verifiable by stakeholders, arbitrators, or insurers in disputes, without accessing underlying data. This ensures three security properties: integrity, authenticity, and validity. Our prototype demonstrates linear scaling up to over 1 million events per hour for measurements with near constant-time proof generation and verification for single violation claims, enabling trustless SLA enforcement through cryptographic guarantees for automated compliance verification in service monitoring.

Despite the rapid progress of neural networks, they remain highly vulnerable to adversarial examples, for which adversarial training (AT) is currently the most effective defense. While AT has been extensively studied, its practical applications expose two major limitations: natural accuracy tends to degrade significantly compared with standard training, and robustness does not transfer well across attacks crafted under different norm constraints. Unlike prior works that attempt to address only one issue within a single network, we propose to partition the overall generalization goal into multiple sub-tasks, each assigned to a dedicated base learner. By specializing in its designated objective, each base learner quickly becomes an expert in its field. In the later stages of training, we interpolate their parameters to form a knowledgeable global learner, while periodically redistributing the global parameters back to the base learners to prevent their optimization trajectories from drifting too far from the shared target. We term this framework Generalist and introduce three variants tailored to different application scenarios. Both theoretical analysis and extensive experiments demonstrate that Generalist achieves lower generalization error and significantly alleviates the trade-off problems compared with baseline methods. Our results suggest that Generalist provides a promising step toward developing fully robust classifiers in the future.

Backdoor attacks pose a persistent security risk to deep neural networks (DNNs) due to their stealth and durability. While recent research has explored leveraging model unlearning mechanisms to enhance backdoor concealment, existing attack strategies still leave persistent traces that may be detected through static analysis. In this work, we introduce the first paradigm of revocable backdoor attacks, where the backdoor can be proactively and thoroughly removed after the attack objective is achieved. We formulate the trigger optimization in revocable backdoor attacks as a bilevel optimization problem: by simulating both backdoor injection and unlearning processes, the trigger generator is optimized to achieve a high attack success rate (ASR) while ensuring that the backdoor can be easily erased through unlearning. To mitigate the optimization conflict between injection and removal objectives, we employ a deterministic partition of poisoning and unlearning samples to reduce sampling-induced variance, and further apply the Projected Conflicting Gradient (PCGrad) technique to resolve the remaining gradient conflicts. Experiments on CIFAR-10 and ImageNet demonstrate that our method maintains ASR comparable to state-of-the-art backdoor attacks, while enabling effective removal of backdoor behavior after unlearning. This work opens a new direction for backdoor attack research and presents new challenges for the security of machine learning systems.

The integration of blockchain technology into healthcare presents a paradigm shift for secure data management, enabling decentralized and tamper-proof storage and sharing of sensitive Electronic Health Records (EHRs). However, existing blockchain-based healthcare systems, while providing robust access control, commonly overlook the high latency in user-side re-computation of hashes for integrity verification of large multimedia data, impairing their practicality, especially in time-sensitive clinical scenarios. In this paper, we propose FAITH, an innovative scheme for \underline{F}ast \underline{A}uthenticated and \underline{I}nteroperable mul\underline{T}imedia \underline{H}ealthcare data storage and sharing over hybrid-storage blockchains. Rather than user-side hash re-computations, FAITH lets an off-chain storage provider generate verifiable proofs using recursive Zero-Knowledge Proofs (ZKPs), while the user only needs to perform lightweight verification. For flexible access authorization, we leverage Proxy Re-Encryption (PRE) and enable the provider to conduct ciphertext re-encryption, in which the re-encryption correctness can be verified via ZKPs against the malicious provider. All metadata and proofs are recorded on-chain for public verification. We provide a comprehensive analysis of FAITH's security regarding data privacy and integrity. We implemented a prototype of FAITH, and extensive experiments demonstrated its practicality for time-critical healthcare applications, dramatically reducing user-side verification latency by up to $98\%$, bringing it from $4$ s down to around $70$ ms for a $5$ GB encrypted file.

Rule-based network intrusion detection systems play a crucial role in the real-time detection of Web attacks. However, most existing works primarily focus on automatically generating detection rules for new attacks, often overlooking the relationships between new attacks and existing rules, which leads to significant redundancy within the ever-expanding ruleset. To address this issue, we propose GRIDAI, a novel end-to-end framework for the automated Generation and Repair of Intrusion Detection rules through collaboration among multiple LLM-based agents. Unlike traditional methods, GRIDAI first assesses the nature of incoming attack samples. If the sample represents a new attack type, it is used to generate a new rule. Otherwise, the sample is identified as a variant of an attack already covered by an existing rule and used to repair the rule by updating the corresponding signature, thereby enhancing its generalization capability. Additionally, to mitigate syntactic and semantic errors in rules caused by LLM hallucinations, we incorporate a tool-based real-time validation mechanism and a representative attack sample maintained for each rule, enabling fully automated rule generation and repair. Comprehensive experiments were conducted on a public dataset containing seven types of attacks and a private dataset with 43 attack types. The results demonstrate that GRIDAI accurately identifies the relationships between new attack samples and existing rules, efficiently generates and repairs rules to handle new attacks and variants, and effectively mitigates the impact of LLM hallucinations.

We report on two months of ethnographic fieldwork in a women's centre in Pattaya, and interviews with 76 participants. Our findings, as they relate to digital security, show how (i) women in Pattaya, often working in the sex and massage industries, perceived relationships with farang men as their best, and sometimes only, option to achieve security; (ii) the strategies used by the women to appeal to a farang involved presenting themselves online, mirroring how they were being advertised by bar owners to attract customers; (iii) appealing to what they considered `Western ideals', the women sought out `Western technologies' and appropriated them for their benefit; (iv) the women navigated a series of online security risks, such as scams and abuse, which shaped their search for a farang; (v) the women developed collective security through knowledge-sharing to protect themselves and each other in their search for a farang. We situate our work in emerging digital security scholarship within marginalised contexts.

Indoor robotic systems within Cyber-Physical Systems (CPS) are increasingly exposed to Denial of Service (DoS) attacks that compromise localization, control and telemetry integrity. We propose a privacy-aware malware detection framework for indoor robotic systems, which leverages hybrid quantum computing and deep neural networks to counter DoS threats in CPS, while preserving privacy information. By integrating quantum-enhanced feature encoding with dropout-optimized deep learning, our architecture achieves up to 95.2% detection accuracy under privacy-constrained conditions. The system operates without handcrafted thresholds or persistent beacon data, enabling scalable deployment in adversarial environments. Benchmarking reveals robust generalization, interpretability and resilience against training instability through modular circuit design. This work advances trustworthy AI for secure, autonomous CPS operations.

The run-time electromagnetic (EM) emanation of microprocessors presents a side-channel that leaks the confidentiality of the applications running on them. Many recent works have demonstrated successful attacks leveraging such side-channels to extract the confidentiality of diverse applications, such as the key of cryptographic algorithms and the hyperparameter of neural network models. This paper proposes ShuffleV, a microarchitecture defense strategy against EM Side-Channel Attacks (SCAs). ShuffleV adopts the moving target defense (MTD) philosophy, by integrating hardware units to randomly shuffle the execution order of program instructions and optionally insert dummy instructions, to nullify the statistical observation by attackers across repetitive runs. We build ShuffleV on the open-source RISC-V core and provide six design options, to suit different application scenarios. To enable rapid evaluation, we develop a ShuffleV simulator that can help users to (1) simulate the performance overhead for each design option and (2) generate an execution trace to validate the randomness of execution on their workload. We implement ShuffleV on a Xilinx PYNQ-Z2 FPGA and validate its performance with two representative victim applications against EM SCAs, AES encryption, and neural network inference. The experimental results demonstrate that ShuffleV can provide automatic protection for these applications, without any user intervention or software modification.

Tools focused on cryptographic API misuse often detect the most basic expressions of the vulnerable use, and are unable to detect non-trivial variants. The question of whether tools should be designed to detect such variants can only be answered if we know how developers use and misuse cryptographic APIs in the wild, and in particular, what the unnatural usage of such APIs looks like. This paper presents the first large-scale study that characterizes unnatural crypto-API usage through a qualitative analysis of 5,704 representative API invocations. We develop an intuitive complexity metric to stratify 140,431 crypto-API invocations obtained from 20,508 Android applications, allowing us to sample 5,704 invocations that are representative of all strata, with each stratum consisting of invocations with similar complexity/naturalness. We qualitatively analyze the 5,704 sampled invocations using manual reverse engineering, through an in-depth investigation that involves the development of minimal examples and exploration of native code. Our study results in two detailed taxonomies of unnatural crypto-API misuse, along with 17 key findings that show the presence of highly unusual misuse, evasive code, and the inability of popular tools to reason about even mildly unconventional usage. Our findings lead to four key takeaways that inform future work focused on detecting unnatural crypto-API misuse.

Within the cyber-physical-social-climate nexus, all systems are deeply interdependent: cyber infrastructure facilitates communication, data processing, and automation across physical systems (such as power grids and networks), while social infrastructure provides the human capital and societal norms necessary for the system's functionality. Any disruption within any of these components, whether due to human error or system mismanagement, can propagate throughout the network, amplifying vulnerabilities and creating a significantly scaled impact. This chapter explores the critical role of human vulnerabilities within the cyber-physical-social-climate nexus, focusing on the interdependencies across cyber, physical, and social systems and how these interdependencies can scale in a climate context. While cyber and physical vulnerabilities are readily apparent, social vulnerabilities (such as misinformation, resistance to policy change, and lack of public awareness) often go unaddressed despite their profound impact on resilience and climate adaptation. Social infrastructure, including human capital, societal norms, and policy frameworks, shapes community responses and underpins adaptive capacity, yet it is also a significant point of failure when overlooked. This chapter examines how human cognitive biases, risk misperception, and decision-making silos within interconnected systems can lead to resource misallocation and weakened policy effectiveness. These factors are analyzed to demonstrate how inadequate responses across cyber-physical-social layers can cascade, amplifying climate-related risks. By addressing these human factors and aligning decision-making frameworks, we aim to strengthen resilience and foster cohesive adaptation strategies that account for the intricate interrelations of cyber-physical-social-climate systems.

Analyzing 500 CTF participants, this paper shows that while participants readily bypassed simple AI guardrails using common techniques, layered multi-step defenses still posed significant challenges, offering concrete insights for building safer AI systems.

The Model Context Protocol (MCP) standardizes how large language model (LLM) agents discover, describe, and call external tools. While MCP unlocks broad interoperability, it also enlarges the attack surface by making tools first-class, composable objects with natural-language metadata, and standardized I/O. We present MSB (MCP Security Benchmark), the first end-to-end evaluation suite that systematically measures how well LLM agents resist MCP-specific attacks throughout the full tool-use pipeline: task planning, tool invocation, and response handling. MSB contributes: (1) a taxonomy of 12 attacks including name-collision, preference manipulation, prompt injections embedded in tool descriptions, out-of-scope parameter requests, user-impersonating responses, false-error escalation, tool-transfer, retrieval injection, and mixed attacks; (2) an evaluation harness that executes attacks by running real tools (both benign and malicious) via MCP rather than simulation; and (3) a robustness metric that quantifies the trade-off between security and performance: Net Resilient Performance (NRP). We evaluate nine popular LLM agents across 10 domains and 405 tools, producing 2,000 attack instances. Results reveal the effectiveness of attacks against each stage of MCP. Models with stronger performance are more vulnerable to attacks due to their outstanding tool calling and instruction following capabilities. MSB provides a practical baseline for researchers and practitioners to study, compare, and harden MCP agents. Code: https://github.com/dongsenzhang/MSB

Federated learning (FL) enables organizations to collaboratively train models without sharing their datasets. Despite this advantage, recent studies show that both client updates and the global model can leak private information, limiting adoption in sensitive domains such as healthcare. Local differential privacy (LDP) offers strong protection by letting each participant privatize updates before transmission. However, existing LDP methods were designed for centralized training and introduce challenges in FL, including high resource demands that can cause client dropouts and the lack of reliable privacy guarantees under asynchronous participation. These issues undermine model generalizability, fairness, and compliance with regulations such as HIPAA and GDPR. To address them, we propose L-RDP, a DP method designed for LDP that ensures constant, lower memory usage to reduce dropouts and provides rigorous per-client privacy guarantees by accounting for intermittent participation.

Modern web and digital application password storage relies on password hashing for storage and security. Ad-hoc upgrade of password storage to keep up with hash algorithm norms may be used to save costs but can introduce unforeseen vulnerabilities. This is the case in the password storage scheme used by Meta Platforms which services several billion monthly users worldwide. In this paper we present the first example of an exploit which demonstrates the security weakness of Facebook's password storage scheme, and discuss its implications. Proper ethical disclosure guidelines and vendor notification were followed.

As e-commerce platforms develop, fraudulent activities are increasingly emerging, posing significant threats to the security and stability of these platforms. Promotion abuse is one of the fastest-growing types of fraud in recent years and is characterized by users exploiting promotional activities to gain financial benefits from the platform. To investigate this issue, we conduct the first study on promotion abuse fraud in e-commerce platforms MEITUAN. We find that promotion abuse fraud is a group-based fraudulent activity with two types of fraudulent activities: Stocking Up and Cashback Abuse. Unlike traditional fraudulent activities such as fake reviews, promotion abuse fraud typically involves ordinary customers conducting legitimate transactions and these two types of fraudulent activities are often intertwined. To address this issue, we propose leveraging additional information from the spatial and temporal perspectives to detect promotion abuse fraud. In this paper, we introduce PROMOGUARDIAN, a novel multi-relation fused graph neural network that integrates the spatial and temporal information of transaction data into a homogeneous graph to detect promotion abuse fraud. We conduct extensive experiments on real-world data from MEITUAN, and the results demonstrate that our proposed model outperforms state-of-the-art methods in promotion abuse fraud detection, achieving 93.15% precision, detecting 2.1 to 5.0 times more fraudsters, and preventing 1.5 to 8.8 times more financial losses in production environments.

In modern containerized cloud environments, the adoption of RDMA (Remote Direct Memory Access) has expanded to reduce CPU overhead and enable high-performance data exchange. Achieving this requires strong performance isolation to ensure that one container's RDMA workload does not degrade the performance of others, thereby maintaining critical security assurances. However, existing isolation techniques are difficult to apply effectively due to the complexity of microarchitectural resource management within RDMA NICs (RNICs). This paper experimentally analyzes two types of resource exhaustion attacks on NVIDIA BlueField-3: (i) state saturation attacks and (ii) pipeline saturation attacks. Our results show that state saturation attacks can cause up to a 93.9% loss in bandwidth, a 1,117x increase in latency, and a 115% rise in cache misses for victim containers, while pipeline saturation attacks lead to severe link-level congestion and significant amplification, where small verb requests result in disproportionately high resource consumption. To mitigate these threats and restore predictable security assurances, we propose HT-Verbs, a threshold-driven framework based on real-time per-container RDMA verb telemetry and adaptive resource classification that partitions RNIC resources into hot, warm, and cold tiers and throttles abusive workloads without requiring hardware modifications.

Unclonable cryptography uses the quantum no-cloning principle to achieve security guarantees that are impossible in a classical world. Most existing works, however, consider only single-key security. The few works that achieve collusion-resistance rely on techniques highly tailored to their applications. Moreover, multi-copy security, where an adversary receives identical copies of a pure state, remains largely open. In this work, we develop a toolset of generic compilers and technical lemmata for collusion-resistance and multi-copy security. We first show how classical functional encryption can upgrade single-key quantum protection of decryption keys to collusion-resistant security, covering copy-protection, secure leasing, and leakage resilience. We then give a purification compiler that upgrades collusion-resistant state-query games with classically determined outputs to multi-copy security, assuming only one-way functions. Along the way, we develop a collusion-resistant one-way-to-hiding lemma, a quantum-state analogue of the small-range-distributions lemma, and a quantum pigeonhole lemma for entangled adversaries. We also construct the first deterministic signatures with quantum-query security and show that strong search security implies identical-challenge security for collusion-resistant single-decryptor encryption. Using these tools, we obtain the first public-key quantum coins, multi-copy secure single-decryptor encryption and unclonable encryption, and collusion-resistant secure key leasing with a fully classical vendor. Our quantum coins and plain-model search-secure UE use indistinguishability obfuscation and one-way functions, while our QROM UE achieves indistinguishability security unconditionally. Finally, we obtain multi-copy secure LOCC leakage-resilient public-key encryption, making progress towards quantum key-fire in the plain model.

Confidential Virtual Machines (CVMs) protect data in use by running workloads within hardware-enforced Trusted Execution Environments (TEEs). However, existing CVM attestation mechanisms only certify what code is running, not where it is running. Commercial TEEs mitigate passive physical attacks through memory encryption but explicitly exclude active hardware tampering (memory interposers, physical side channels, ...). Yet current attestations provide no cryptographic evidence that a CVM executes on hardware residing within a trusted data center where such attacks would not take place. This gap enables proxy attacks in which valid attestations are combined across machines to falsely attest trusted execution. To bridge this gap, we introduce Data Center Execution Assurance (DCEA), a design that generates a cryptographic Proof of Cloud by binding CVM attestation to platform-level Trusted Platform Module (TPM) evidence. DCEA combines two independent roots of trust. First, the TEE manufacturer, and second, the infrastructure provider, by cross-linking runtime TEE measurements with the vTPM-measured boot CVM state. This binding ensures that CVM execution, vTPM quotes, and platform provenance all originate from the same physical chassis. We formalize the environment's provenance and show that DCEA prevents advanced relay attacks, including a novel mix-and-match proxy attack. Using the AGATE framework in the Universal Composability model, we prove that DCEA emulates an ideal location-aware TEE even under a malicious host software stack. We implement DCEA on Google Cloud bare-metal Intel TDX instances using Intel TXT and evaluate its performance, demonstrating practical overheads and deployability. DCEA refines the CVM threat model and enables verifiable execution-location guarantees for privacy-sensitive workloads.

Large Language Models (LLMs) are increasingly being used to autonomously evaluate the quality of content in communication systems, e.g., to assess responses in telecom customer support chatbots. However, the impartiality of these AI "judges" is not guaranteed, and any biases in their evaluation criteria could skew outcomes and undermine user trust. In this paper, we systematically investigate judgment biases across 6 LLM-as-a-judge models spanning both prompt-based and fine-tuned judges under the pointwise scoring setting, encompassing 11 types of biases that cover both implicit and explicit forms. We observed that state-of-the-art LLM judges demonstrate robustness to biased inputs, generally assigning them lower scores than the corresponding clean samples. We further found that fine-tuning an LLM on high-scoring yet biased responses can significantly degrade its performance, highlighting the risk of training on biased data. We also discovered that the judged scores correlate with task difficulty: a challenging dataset like GPQA yields lower average scores, whereas an open-ended reasoning dataset (e.g., JudgeLM-val) sees higher average scores. Finally, we proposed four potential mitigation strategies to ensure fair and reliable AI judging in practical communication scenarios.

The growing scale and sophistication of cyberattacks pose critical challenges to network security, particularly in detecting diverse intrusion types within imbalanced datasets. Traditional intrusion detection systems (IDS) often struggle to maintain high accuracy across both frequent and rare attacks, leading to increased false negatives for minority classes. To address this, we propose a hybrid anomaly detection framework that integrates specialized deep learning models with an ensemble meta-classifier. Each model is trained to detect a specific attack category, enabling tailored learning of class-specific patterns, while their collective outputs are fused by a Random Forest meta-classifier to improve overall decision reliability. The framework is evaluated on the NSL-KDD benchmark, demonstrating superior performance in handling class imbalance compared to conventional monolithic models. Results show significant improvements in precision, recall, and F1-score across all attack categories, including rare classes such as User to Root (U2R). The proposed system achieves near-perfect detection rates with minimal false alarms, highlighting its robustness and generalizability. This work advances the design of intrusion detection systems by combining specialization with ensemble learning, providing an effective and scalable solution for safeguarding modern networks.

Traditional security scanners fail when facing new attack patterns they haven't seen before. They rely on fixed rules and predetermined signatures, making them blind to novel threats. We present a fundamentally different approach: instead of memorizing specific attack patterns, we learn what makes systems genuinely secure. Our key insight is simple yet powerful: context determines vulnerability. A SQL query that's safe in one environment becomes dangerous in another. By modeling this context-vulnerability relationship, we achieve something remarkable: our system detects attacks it has never seen before. We introduce context-aware verification that learns from genuine system behavior. Through reconstruction learning on secure systems, we capture their essential characteristics. When an unknown attack deviates from these patterns, our system recognizes it, even without prior knowledge of that specific attack type. We prove this capability theoretically, showing detection rates improve exponentially with context information I(W;C). Our framework combines three components: (1) reconstruction learning that models secure behavior, (2) multi-scale graph reasoning that aggregates contextual clues, and (3) attention mechanisms guided by reconstruction differences. Extensive experiments validate our approach: detection accuracy jumps from 58 percent to 82 percent with full context, unknown attack detection improves by 31 percent, and our system maintains above 90 percent accuracy even against completely novel attack vectors.