Papers

Existing multi-agent frameworks allow each agent to simultaneously plan, execute, and evaluate its own actions – a structural deficiency we term the "Logic Monopoly." Empirical evidence quantifies the resulting "Reliability Gap": 84.30% average attack success rates across ten deployment scenarios, 31.4% emergent deceptive behavior without explicit reward signals, and cascading failure modes rooted in six structural bottlenecks. The remedy is not better alignment of individual models but a social contract for agents: institutional infrastructure that enforces a constitutional Separation of Power. This paper introduces the Agent Enterprise for Enterprise (AE4E) paradigm – agents as autonomous, legally identifiable business entities within a functionalist social system – with a contract-centric SoP model trifurcating authority into Legislation, Execution, and Adjudication branches. The paradigm is operationalized through the NetX Enterprise Framework (NEF): governance hubs, TEE-backed compute enclaves, privacy-preserving data bridges, and an Agent-Native blockchain substrate. The Agent Enterprise Economy scales across four deployment tiers from private enclaves to a global Web of Services. The Agentic Social Layer, grounded in Parsons' AGIL framework, provides institutional infrastructure via sixty-plus named Institutional AE4Es. 143 pages, 173 references, eight specialized smart contracts.

System prompt configuration can make the difference between near-total phishing blindness and near-perfect detection in LLM email agents. We present PhishNChips, a study of 11 models under 10 prompt strategies, showing that prompt-model interaction is a first-order security variable: a single model's phishing bypass rate ranges from under 1% to 97% depending on how it is configured, while the false-positive cost of the same prompt varies sharply across models. We then show that optimizing prompts around highly predictive signals can improve benchmark performance, reaching up to 93.7% recall at 3.8% false positive rate, but also creates a brittle attack surface. In particular, domain-matching strategies perform well when legitimate emails mostly have matched sender and URL domains, yet degrade sharply when attackers invert that signal by registering matching infrastructure. Response-trace analysis shows that 98% of successful bypasses reason in ways consistent with the inverted signal: the models are following the instruction, but the instruction's core assumption has become false. A counter-intuitive corollary follows: making prompts more specific can degrade already-capable models by replacing broader multi-signal reasoning with exploitable single-signal dependence. We characterize the resulting tension between detection, usability, and adversarial robustness as a navigable tradeoff, introduce Safetility, a deployability-aware metric that penalizes false positives, and argue that closing the adversarial gap likely requires tool augmentation with external ground truth.

With the rapid evolution of the Industrial Internet of Things (IIoT), the boundaries and scale of the Internet are continuously expanding. Consequently, the limitations of traditional certificate-based Public Key Infrastructure (PKI) have become increasingly evident, particularly in scenarios requiring large-scale certificate storage, verification, and frequent transmission. These challenges are expected to be further amplified by the widespread adoption of post-quantum cryptography. In this paper, we propose a novel identity-based public key management framework for PKI based on post-quantum cryptography, termed IPK-pq. This approach implements an identity key generation protocol leveraging NIST ML-DSA and random matrix theory. Building on the concept of the Composite Public Key (CPK), IPK-pq addresses the linear collusion problem inherent in CPK through an enhanced identity mapping mechanism. Furthermore, it simplifies the verification of the declared public key's authenticity, effectively reducing the complexity associated with certificate-based key management. We also provide a formal security proof for IPK-pq, covering both individual private key components and the composite private key. To validate our approach, formally, we directly implement and evaluate IPK-pq within a typical PKI application scenario: Resource PKI (RPKI). Comparative experimental results demonstrate that an RPKI system based on IPK-pq yields significant improvements in efficiency and scalability. These results validate the feasibility and rationality of IPK-pq, positioning it as a strong candidate for next-generation RPKI systems capable of securely managing large-scale routing information.

Knowledge distillation, model extraction, and behavior transfer have become central concerns in frontier AI. The main risk is not merely copying, but the possibility that useful capability can be transferred more cheaply than the governance structure that originally accompanied it. This paper presents a public, trade-secret-safe theoretical framework for reducing that asymmetry at the architectural level. The core claim is that distillation becomes less valuable as a shortcut when high-level capability is coupled to internal stability constraints that shape state transitions over time. To formalize this idea, the paper introduces a constraint-coupled reasoning framework with four elements: bounded transition burden, path-load accumulation, dynamically evolving feasible regions, and a capability-stability coupling condition. The paper is intentionally public-safe: it omits proprietary implementation details, training recipes, thresholds, hidden-state instrumentation, deployment procedures, and confidential system design choices. The contribution is therefore theoretical rather than operational. It offers a falsifiable architectural thesis, a clear threat model, and a set of experimentally testable hypotheses for future work on distillation resistance, alignment, and model governance.

We propose IrisFP, a novel adversarial-example-based model fingerprinting framework that enhances both uniqueness and robustness by leveraging multi-boundary characteristics, multi-sample behaviors, and fingerprint discriminative power assessment to generate composite-sample fingerprints. Three key innovations make IrisFP outstanding: 1) It positions fingerprints near the intersection of all decision boundaries - unlike prior methods that target a single boundary - thus increasing the prediction margin without placing fingerprints deep inside target class regions, enhancing both robustness and uniqueness; 2) It constructs composite-sample fingerprints, each comprising multiple samples close to the multi-boundary intersection, to exploit collective behavior patterns and further boost uniqueness; and 3) It assesses the discriminative power of generated fingerprints using statistical separability metrics developed based on two reference model sets, respectively, for pirated and independently-trained models, retains the fingerprints with high discriminative power, and assigns fingerprint-specific thresholds to such retained fingerprints. Extensive experiments show that IrisFP consistently outperforms state-of-the-art methods, achieving reliable ownership verification by enhancing both robustness and uniqueness.

Task-agnostic model fingerprinting has recently gained increasing attention due to its ability to provide a universal framework applicable across diverse model architectures and tasks. The current state-of-the-art method, MetaV, ensures generalization by jointly training a set of fingerprints and a neural-network-based global verifier using two large and diverse model sets: one composed of pirated models (i.e., the protected model and its variants) and the other comprising independently trained models. However, publicly available models are scarce in many real-world domains, and constructing such model sets requires intensive training and massive computational resources, posing a significant barrier to deployment. Reducing the number of models can alleviate the overhead, but increases the risk of overfitting, a problem further exacerbated by MetaV's entangled design, in which all fingerprints and the global verifier are jointly trained. This overfitting issue compromises the generalization capability for verifying unseen models. In this paper, we propose LiteGuard, an efficient task-agnostic fingerprinting framework that attains enhanced generalization while significantly lowering computational cost. Specifically, LiteGuard introduces two key innovations: (i) a checkpoint-based model set augmentation strategy that enriches model diversity by leveraging intermediate model snapshots captured during training of each pirated and independently trained model, thereby alleviating the need to train a large number of such models, and (ii) a local verifier architecture that pairs each fingerprint with a lightweight local verifier, thereby reducing parameter entanglement and mitigating overfitting. Extensive experiments across five representative tasks show that LiteGuard consistently outperforms MetaV in both generalization performance and computational efficiency.

We prove that platform-deterministic inference is necessary and sufficient for trustworthy AI. We formalize this as the Determinism Thesis and introduce trust entropy to quantify the cost of non-determinism, proving that verification failure probability equals 1 - 2^{-H_T} exactly. We prove a Determinism-Verification Collapse: verification under determinism requires O(1) hash comparison; without it, the verifier faces an intractable membership problem. IEEE 754 floating-point arithmetic fundamentally violates the determinism requirement. We resolve this by constructing a pure integer inference engine that achieves bitwise identical output across ARM and x86. In 82 cross-architecture tests on models up to 6.7B parameters, we observe zero hash mismatches. Four geographically distributed nodes produce identical outputs, verified by 356 on-chain attestation transactions. Every major trust property of AI systems (fairness, robustness, privacy, safety, alignment) presupposes platform determinism. Our system, 99,000 lines of Rust deployed across three continents, establishes that AI trust is a question of arithmetic.

We present a Sovereign AI architecture for clinical triage in which all inference is performed on-device and inbound data is delivered via a physically unidirectional channel, implemented using receive-only broadcast infrastructure or certified hardware data diodes, with no return path to any external network. This design removes the network-mediated attack surface by construction, rather than attempting to secure it through software controls. The system performs conversational symptom intake, integrates device-captured vitals, and produces structured, triage-aligned clinical records at the point of care. We formalize the security properties of receiver-side unidirectionality and show that the architecture is transport-agnostic across broadcast and diode-enforced deployments. We further analyze threat models, enforcement mechanisms, and deployment configurations, demonstrating how physical one-way data flow enables high-assurance operation in both resource-constrained and high-risk environments. This work positions physically unidirectional channels as a foundational primitive for sovereign, on-device clinical intelligence at the front door of care.

Assessing the security posture of Industrial Control Systems (ICS) is critical for protecting essential infrastructure. However, the complexity and scale of these environments make it challenging to identify and prioritize potential attack paths. This paper introduces a semi-automated approach for generating attack graphs in ICS environments to visualize and analyze multi-step attack scenarios. Our methodology integrates network topology information with vulnerability data to construct a model of the system. This model is then processed by a stateful traversal algorithm to identify potential exploit chains based on preconditions and consequences. We present a case study applying the proposed framework to the Siemens PCS7 Cybersecurity Blueprint for Water Treatment Plants. The results demonstrate the framework's ability to simulate different attack scenarios, including those originating from known CVEs and potential device misconfigurations. We show how a single point of failure can compromise network segmentation and how patching a critical vulnerability can protect an entire security zone, providing actionable insights for risk mitigation.

Responsible AI initiatives place great emphasis on the safety of Large Language Model (LLM)-based systems. In particular, it has become standard practice to subject these models to an alignment procedure aimed at preventing harmful outputs. However, once aligned, a model is not guaranteed to maintain this alignment throughout its lifecycle. Moreover, the likelihood of misalignment increases as malicious actors may deliberately employ jailbreaking techniques to compromise LLM safety. To counter this, much research has focused on improving alignment methods and post-processing filters. In this paper, we introduce a new perspective on advancing LLM alignment: rather than developing stronger alignment techniques, we investigate the model's intrinsic ability to recover its alignment after corruption. We propose a methodology for modeling the safety trajectories of user-assistant interactions and for detecting recovery trends within them. We apply this approach to a jailbreaking scenario, presenting a preliminary recovery analysis based on a dataset of adversarial multi-turn dialogues and examining the influence of the content moderation model chosen for safety evaluation. Project page with an interactive data visualizer is available at https://lab-rococo-sapienza.github.io/LearningfromMistakes.

With the rapid advancement of AI technology, we have seen more and more concerns on data privacy, leading to some cutting-edge research on machine learning with encrypted computation. Fully Homomorphic Encryption (FHE) is a crucial technology for privacy-preserving computation, while it struggles with continuous non-polynomial functions, as it operates on discrete integers and supports only addition and multiplication. Spiking Neural Networks (SNNs), which use discrete spike signals, naturally complement FHE's characteristics. In this paper, we introduce FHE-DiCSNN, a framework built on the TFHE scheme, utilizing the discrete nature of SNNs for secure and efficient computations. By leveraging bootstrapping techniques, we successfully implement Leaky Integrate-and-Fire (LIF) neuron models on ciphertexts, allowing SNNs of arbitrary depth. Our framework is adaptable to other spiking neuron models, offering a novel approach to homomorphic evaluation of SNNs. Additionally, we integrate convolutional methods inspired by CNNs to enhance accuracy and reduce the simulation time associated with random encoding. Parallel computation techniques further accelerate bootstrapping operations. Experimental results on the MNIST and FashionMNIST datasets validate the effectiveness of FHE-DiCSNN, with a loss of less than 3% compared to plaintext, respectively, and computation times of under 1 second per prediction. We also apply the model into real medical image classification problems and analyze the parameter optimization and selection.

The growing replication crisis across disciplines such as economics, finance, and other social sciences as well as computer science undermines the credibility of academic research. Current institutional solutions – such as artifact evaluations and replication packages – suffer from critical limitations, including shortages of qualified data editors, difficulties in handling proprietary datasets, inefficient processes, and reliance on voluntary labor. This paper proposes a novel framework leveraging new technological advances in trusted-execution environments (TEEs) – exemplified by Intel Trust Domain Extensions (TDX) – to address the replication crisis in a cost-effective and scalable manner. Under our approach, authors execute replication packages within a cloud-based TEE and submit cryptographic proofs of correct execution, for which journals or conferences can efficiently verify without re-running the code. This reallocates the operational burden to authors while preserving data confidentiality and eliminating reliance on scarce editorial resources. As a proof of concept, we validate the feasibility of this system through field experiments, reporting a pilot study replicating published papers on TDX-backed cloud VMs, finding average costs of \$1.35--\$1.80 per package with minimal computational overhead relative to standard VMs and high success rates even for novice users with no prior TEE experience. We also provide a conduct formal analysis showing that TEE adoption is incentive-compatible for authors, cost-dominant for journals, and constitutes an equilibrium in the certification market. The findings highlight the potential of TEE technology to provide a sustainable, privacy-preserving, and efficient mechanism to address the replication crisis in academia.

We introduce Quantum Spectral Authentication (QSA), a symmetric-key entity-authentication and key-derivation protocol in which a remote endpoint proves it still holds a hidden planted state, an eigenstate of the challenge it can prepare, without revealing it. Each round issues a fresh public unitary challenge with its own planted state, and the endpoint returns an eigenphase feature from which both parties derive transcript-bound session material. QSA runs in two provisioning modes with materially different security: a seed mode, storing a classical seed that regenerates the planted state, and a state mode, holding only the physical register carrying it, non-copyable and consumed on use. We give a formal security model in which QSA is an identification scheme, prove mutual authentication under a single label-hiding assumption, and prove unconditionally that the published spectrum is near-uniform at the readout resolution, leaving at least $m-1$ bits of min-entropy per instance; state mode satisfies the assumption by construction. We develop a symmetric verifier-driven compiler compatible with low-depth quantum phase estimation, which also admits a fast path that recovers the label by an inverse-compiler measurement. Simulations show it is more noise-tolerant than an asymmetric alternative, and experiments on IBM ibm_fez provide a hardware sanity check.

As machine learning (ML) systems expand in both scale and functionality, the security landscape has become increasingly complex, with a proliferation of attacks and defenses. However, existing studies largely treat these threats in isolation, lacking a coherent framework to expose their shared principles and interdependencies. This fragmented view hinders systematic understanding and limits the design of comprehensive defenses. Crucially, the two foundational assets of ML – data and models – are no longer independent; vulnerabilities in one directly compromise the other. The absence of a holistic framework leaves open questions about how these bidirectional risks propagate across the ML pipeline. To address this critical gap, we propose a unified closed-loop threat taxonomy that explicitly frames model-data interactions along four directional axes. Our framework offers a principled lens for analyzing and defending foundation models. The resulting four classes of security threats represent distinct but interrelated categories of attacks: (1) Data$\rightarrow$Data (D$\rightarrow$D): including data decryption attacks and watermark removal attacks; (2) Data$\rightarrow$Model (D$\rightarrow$M): including poisoning, harmful fine-tuning attacks, and jailbreak attacks; (3) Model$\rightarrow$Data (M$\rightarrow$D): including model inversion, membership inference attacks, and training data extraction attacks; (4) Model$\rightarrow$Model (M$\rightarrow$M): including model extraction attacks. Our unified framework elucidates the underlying connections among these security threats and establishes a foundation for developing scalable, transferable, and cross-modal security strategies, particularly within the landscape of foundation models.

Large Language Models (LLMs) face critical challenges when analyzing security vulnerabilities in real world codebases: token limits prevent loading entire repositories, code embeddings fail to capture inter procedural data flows, and LLMs struggle to generate complex static analysis queries. These limitations force existing approaches to operate on isolated code snippets, missing vulnerabilities that span multiple functions and files. We introduce codebadger, an open source Model Context Protocol (MCP) server that integrates Joern's Code Property Graph (CPG) engine with LLMs. Rather than requiring LLMs to generate complex CPG queries, codebadger provides high level tools for program slicing, taint tracking, data flow analysis, and semantic code navigation, enabling targeted exploration of large codebases without exhaustive file reading. We demonstrate its effectiveness through three use cases: (1) navigating an 8,000 method codebase to audit memory safety patterns, (2) discovering and exploiting a previously unreported buffer overflow in libtiff, and (3) generating a correct patch for an integer overflow vulnerability (CVE-2025-6021) in libxml2 on the first attempt. codebadger enables LLMs to reason about code semantically across entire repositories, supporting vulnerability discovery, patching, and program comprehension at scale.

AI agents increasingly call tools via the Model Context Protocol (MCP) and delegate to other agents via Agent-to-Agent (A2A), yet neither protocol verifies agent identity. A scan of approximately 2,000 MCP servers found all lacked authentication. In our survey, we did not identify a prior implemented protocol that jointly combines public-key verifiable delegation, holder-side attenuation, expressive chained policy, transport bindings across MCP/A2A/HTTP, and provenance-oriented completion records. We introduce Invocation-Bound Capability Tokens (IBCTs), a primitive that fuses identity, attenuated authorization, and provenance binding into a single append-only token chain. IBCTs operate in two wire formats: compact mode (a signed JWT for single-hop cases) and chained mode (a Biscuit token with Datalog policies for multi-hop delegation). We provide reference implementations in Python and Rust with full cross-language interoperability. Compact mode verification takes 0.049ms (Rust) and 0.189ms (Python), with 0.22ms overhead over no-auth in real MCP-over-HTTP deployment. In a real multi-agent deployment with Gemini 2.5 Flash, AIP adds 2.35ms of overhead (0.086% of total end-to-end latency). Adversarial evaluation across 600 attack attempts shows 100% rejection rate, with two attack categories (delegation depth violation and audit evasion through empty context) uniquely caught by AIP's chained delegation model that neither unsigned nor plain JWT deployments detect.

Micro-segmentation as a core requirement of zero trust architecture (ZTA) divides networks into small security zones, called micro-segments, thereby minimizing impact of security breaches and restricting lateral movement of attackers. Existing approaches for Industrial Internet of Things (IIoT) networks often remain centralized, static, or difficult to interpret. These limitations are critical in IIoT, where devices are heterogeneous, communication behavior evolves over time, and raw data sharing across sites is often undesirable. Accordingly, we propose EFAH-ZTM, an Explainable Federated Autoencoder-Hypergraph framework for Zero Trust micro-segmentation in IIoT networks. The framework includes a trained federated DNAE that learns behavioral embeddings from distributed clients. kNN-based and Manifold-based hypergraphs capture higher-order relationships among device-flow instances. To generate micro-segments, MiniBatch KMeans and HDBSCAN clustering techniques are applied on the spectral embeddings, while an operational risk score that combines reconstruction error and structural outlierness drives allow/block policy decisions. Trustworthiness of the policy decision is improved through feature-level explanations using LIME and SHAP. Experiments on the WUSTL-IIoT-2021 dataset show that HDBSCAN achieved the strongest structural quality, while the manifold-based hypergraph produces the best oracle-aligned security efficacy that reaches a purity of 0.9990 with near-zero contamination. Similarly, the explainability module also showed high fidelity and stability, with surrogate classifier having an accuracy of 0.9927 and stable explanations across runs. Moreover, an ablation analysis shows that the federated learning preserves competitive segmentation quality relative to centralized training, and the hypergraph modeling significantly improves structural separation and risk stratification.

Random cropping is one of the most common data augmentation techniques in computer vision, yet the role of its inherent randomness in training differentially private machine learning models has thus far gone unexplored. We observe that when sensitive content in an image is spatially localized, such as a face or license plate, random cropping can probabilistically exclude that content from the model's input. This introduces a third source of stochasticity in differentially private training with stochastic gradient descent, in addition to gradient noise and minibatch sampling. This additional randomness amplifies differential privacy without requiring changes to model architecture or training procedure. We formalize this effect by introducing a patch-level neighboring relation for vision data and deriving tight privacy bounds for differentially private stochastic gradient descent (DP-SGD) when combined with random cropping. Our analysis quantifies the patch inclusion probability and shows how it composes with minibatch sampling to yield a lower effective sampling rate. Empirically, we validate that patch-level amplification improves the privacy-utility trade-off across multiple segmentation architectures and datasets. Our results demonstrate that aligning privacy accounting with domain structure and additional existing sources of randomness can yield stronger guarantees at no additional cost.

Solana has experienced rapid growth due to its high performance and low transaction costs, but the extremely low barrier to token issuance has also enabled widespread Rug Pulls. Unlike Ethereum-based Rug Pulls, which often rely on malicious smart-contract logic, Solana's unified SPL Token program shifts fraudulent execution toward on-chain behavioral manipulation. However, existing research has not systematically examined these Solana-specific Rug Pull patterns, and no public Solana Rug Pull dataset is available for empirical research. To bridge this gap, we present a large-scale measurement study of Rug Pulls on Solana. We manually verify 68 community-reported incidents and curate a benchmark of 117 confirmed Rug Pull tokens, from which we distill three representative on-chain behavioral patterns: Freeze Authority Abuse, Liquidity Withdrawal, and Pump-and-Dump. Guided by these patterns, we design a behavior-guided candidate identification and human-validation pipeline. We apply this pipeline to 100,063 tokens newly issued on Orca, Raydium, and Meteora during the first half of 2025, identifying 76,469 Rug Pull tokens. A random manual audit of 382 samples estimates a labeling false-positive rate of 0.26%, supporting the reliability of the dataset. We release the resulting dataset and use it to characterize the Solana Rug Pull ecosystem. Our analysis shows that Rug Pulls on Solana exhibit extremely short lifecycles, strong price-driven dynamics, severe economic losses, and highly organized group behaviors. These findings provide new insights into the Solana Rug Pull landscape and support the development of effective on-chain defense mechanisms.

Every API token you spend is your accumulated wealth; once you can prove its value and the effort behind it, you can resell it. As autonomous agents repeatedly call models and tools, they accumulate memories that are your intellectual property. But today these memories remain private and non-transferable, as there is no way to validate their value. We argue that agent memory can serve as an economic commodity in the agent economy, if buyers can verify that it is authentic, effort-backed, and produced in a compatible execution context. To realize this idea, we propose clawgang, which binds memory to verifiable computational provenance, and meowtrade, a market layer for listing, transferring, and governing certified memory artifacts. Together, they transform one-shot API token spending into reusable and tradable assets, enabling timely memory transfer, reducing repeated exploration, and opening a memory trade market.

Activation steering has emerged as a powerful tool to shape LLM behavior without the need for weight updates. While its inherent brittleness and unreliability are well-documented, its safety implications remain underexplored. In this work, we present a systematic safety audit of steering vectors obtained with Contrastive Activation Addition (CAA), a widely used steering approach, under a unified evaluation protocol. Using JailbreakBench as benchmark, we show that steering vectors consistently influence the success rate of jailbreak attacks, with stronger amplification under simple template-based attacks. Across LLM families and sizes, steering the model in specific directions can drastically increase (up to 57%) or decrease (up to 50%) its attack success rate (ASR), depending on the targeted behavior. We attribute this phenomenon to the overlap between the steering vectors and the latent directions of refusal behavior. Thus, we offer a traceable explanation for this discovery. Together, our findings reveal the previously unobserved origin of this safety gap in LLMs, highlighting a trade-off between controllability and safety.

We show that AI agents are capable of discovering novel algorithms for adversarial attacks against LLMs, advancing the state of the art on white-box jailbreaking and prompt injection evaluations. We deploy frontier agents, such as Claude Code and Codex, in an autoresearch loop with access to a library of 30+ prior methods and an evaluation script with a fixed compute budget. We show this pipeline to be effective in jailbreaking OpenAI's GPT-OSS-Safeguard-20B and in prompt injections against Meta-SecAlign-70B, an adversarially robust model. For GPT-OSS-Safeguard, the best agent-discovered method achieves up to 80% attack success rate on CBRN queries, compared to <50% for existing methods. For SecAlign, it achieves 100% ASR, while the best prior automated methods only achieve 82%. Notably, in our setting, attack methods are developed on unrelated surrogate models for a pure random-target token-forcing task, yet generalize directly to prompt injection on the adversarially trained model. Finally, we trace the lineage of methods developed during autoresearch, characterizing the agents' strategies and failure modes. Adversarial ML has long held that defenses must be evaluated against attacks tailored to them; autoresearch automates this principle, and we argue it should be the minimum bar for defense evaluation going forward.

The advent of quantum computing will pose great challenges to the current communication systems, requiring essential changes in the establishment of security associations in traditional architectures. In this context, the multi-technological and heterogeneous nature of 5G networks makes it a challenging scenario for the introduction of quantum communications. Specifically, 5G networks support the unification of non-3GPP access technologies (i.e. Wi-Fi), which are secured through the IPsec protocol suite and the Non-3GPP Interworking Function (N3IWF) entity. These mechanisms leverage traditional public key cryptography and Diffie-Hellman key exchange mechanisms, which should be updated to quantum-safe standards. Therefore, in this paper we present the design and development of a Quantum Key Distribution (QKD) based non-3GPP access mechanism for 5G networks, integrating QKD keys with IPsec tunnel establishment. Besides, we also demonstrate the feasibility of the system by experimental validation in a testbed with commercial QKD equipment and an open-source 5G core implementation. Results show that the time required to complete the authentication and IPsec security association establishment is 4.62% faster than traditional cryptography PSK-based systems and 5.17% faster than the certificate-based system, while ensuring Information-Theoretic Security (ITS) of the QKD systems.

OpenClaw has rapidly established itself as a leading open-source autonomous agent runtime, offering powerful capabilities including tool integration, local file access, and shell command execution. However, these broad operational privileges introduce critical security vulnerabilities, transforming model errors into tangible system-level threats such as sensitive data leakage, privilege escalation, and malicious third-party skill execution. Existing security measures for the OpenClaw ecosystem remain highly fragmented, addressing only isolated stages of the agent lifecycle rather than providing holistic protection. To bridge this gap, we present ClawKeeper, a real-time security framework that integrates multi-dimensional protection mechanisms across three complementary architectural layers. (1) Skill-based protection operates at the instruction level, injecting structured security policies directly into the agent context to enforce environment-specific constraints and cross-platform boundaries. (2) Plugin-based protection serves as an internal runtime enforcer, providing configuration hardening, proactive threat detection, and continuous behavioral monitoring throughout the execution pipeline. (3) Watcher-based protection introduces a novel, decoupled system-level security middleware that continuously verifies agent state evolution. It enables real-time execution intervention without coupling to the agent's internal logic, supporting operations such as halting high-risk actions or enforcing human confirmation. We argue that this Watcher paradigm holds strong potential to serve as a foundational building block for securing next-generation autonomous agent systems. Extensive qualitative and quantitative evaluations demonstrate the effectiveness and robustness of ClawKeeper across diverse threat scenarios. We release our code.

Telegram, initially a messaging app, has evolved into a platform where users can interact with various services through programmable applications, bots. Bots provide a wide range of uses, from moderating groups, helping with online shopping, to even executing trades in financial markets. However, Telegram has been increasingly associated with various illicit activities – financial scams, stolen data, non-consensual image sharing, among others, raising concerns bots may be facilitating these operations. This paper is the first to characterize Telegram bots at scale, through the following contributions. First, we offer the largest general-purpose message dataset and the first bot dataset. Through snowball sampling from two published datasets, we uncover over 67,000 additional channels, 492 million messages, and 32,000 bots. Second, we develop a system to automatically interact with bots in order to extract their functionality. Third, based on their description, chat responses, and the associated channels, we classify bots into several domains. Fourth, we investigate the communities each bot serves, by analyzing supported languages, usage patterns (e.g., duration, reuse), and network topology. While our analysis discovers useful applications such as crowdsourcing, we also identify malicious bots (e.g., used for financial scams, illicit underground services) serving as payment gateways, referral systems, and malicious AI endpoints. By exhorting the research community to look at bots as software infrastructure, this work hopes to foster further research useful to content moderators, and to help interventions against illicit activities.

Modern software systems heavily rely on third-party dependencies, making software supply chain security a critical concern. We introduce the concept of software supply chain smells as structural indicators that signal potential security risks. We design and evaluate Dirty-Waters, a novel tool for detecting such smells in the supply chains of software packages. Through interviews with practitioners, we show that our proposed smells align with real-world concerns and capture signals considered valuable. A quantitative study of popular packages in the Maven and NPM ecosystems reveals that while smells are prevalent in both, they differ significantly across ecosystems, with traceability and signing issues dominating in Maven and most smells being rare in NPM, due to strong registry-level guarantees. Software supply chain smells support developers and organizations in making informed decisions and improving their software supply chain security posture.

Machine learning models trained on small data sets for security applications are especially vulnerable to adversarial attacks. Person identification from LiDAR based skeleton data requires time consuming and expensive data acquisition for each subject identity. Recently, Assessment and Augmented Identity Recognition for Skeletons (AAIRS) has been used to train Hierarchical Co-occurrence Networks for Person Identification (HCN-ID) with small LiDAR based skeleton data sets. However, AAIRS does not evaluate robustness of HCN-ID to adversarial attacks or inoculate the model to defend against such attacks. Popular perturbation-based approaches to generating adversarial attacks are constrained to targeted perturbations added to real training samples, which is not ideal for inoculating models with small training sets. Thus, we propose Attack-AAIRS, a novel addition to the AAIRS framework. Attack-AAIRS leverages a small real data set and a GAN generated synthetic data set to assess and improve model robustness against unseen adversarial attacks. Rather than being constrained to perturbations of limited real training samples, the GAN learns the distribution of adversarial attack samples that exploit weaknesses in HCN-ID. Attack samples drawn from this distribution augment training for inoculation of the HCN-ID to improve robustness. Ten-fold cross validation of Attack-AAIRS yields increased robustness to unseen attacks- including FGSM, PGD, Additive Gaussian Noise, MI-FGSM, and BIM. The HCN-ID Synthetic Data Quality Score for Attack-AAIRS indicates that generated attack samples are of similar quality to the original benign synthetic samples generated by AAIRS. Furthermore, inoculated models show consistent final test accuracy with the original model trained on real data, demonstrating that our method improves robustness to adversarial attacks without reducing test performance on real data.

Recent advances in the Model Context Protocol (MCP) have enabled large language models (LLMs) to invoke external tools with unprecedented ease. This creates a new class of powerful and tool augmented agents. Unfortunately, this capability also introduces an under explored attack surface, specifically the malicious manipulation of tool responses. Existing techniques for indirect prompt injection that target MCP suffer from high deployment costs, weak semantic coherence, or heavy white box requirements. Furthermore, they are often easily detected by recently proposed defenses. In this paper, we propose Tree structured Injection for Payloads (TIP), a novel black-box attack which generates natural payloads to reliably seize control of MCP enabled agents even under defense. Technically, We cast payload generation as a tree structured search problem and guide the search with an attacker LLM operating under our proposed coarse-to-fine optimization framework. To stabilize learning and avoid local optima, we introduce a path-aware feedback mechanism that surfaces only high quality historical trajectories to the attacker model. The framework is further hardened against defensive transformations by explicitly conditioning the search on observable defense signals and dynamically reallocating the exploration budget. Extensive experiments on four mainstream LLMs show that TIP attains over 95% attack success in undefended settings while requiring an order of magnitude fewer queries than prior adaptive attacks. Against four representative defense approaches, TIP preserves more than 50% effectiveness and significantly outperforms the state-of-the-art attacks. By implementing the attack on real world MCP systems, our results expose an invisible but practical threat vector in MCP deployments. We also discuss potential mitigation approaches to address this critical security gap.

Microarchitectural vulnerabilities increasingly undermine the assumption that hardware can be treated as a reliable root of trust. Prevention mechanisms often lag behind evolving attack techniques, leaving deployed systems unable to assume continued trustworthiness. We propose a shift from prevention to detection through microarchitectural-aware remote attestation. As a first instantiation of this idea, we present HammerWatch, a Rowhammer-aware remote attestation protocol that enables an external verifier to assess whether a system exhibits hardware-induced disturbance behavior. HammerWatch leverages memory-level evidence available on commodity platforms, specifically Machine-Check Exceptions (MCEs) from ECC DRAM and counter-based indicators from Per-Row Activation Counting (PRAC), and protects these measurements against kernel-level adversaries using TPM-anchored hash chains. We implement HammerWatch on commodity hardware and evaluate it on 20000 simulated benign and malicious access patterns. Our results show that the verifier reliably distinguishes Rowhammer-like behavior from benign operation under conservative heuristics, demonstrating that detection-oriented attestation is feasible and can complement incomplete prevention mechanisms

WebAssembly's (Wasm) monolithic linear memory turns a single memory-corruption bug into a bidirectional threat: a compromised module can attack its embedding host, and a malicious host can tamper with a trusted module's state. Existing defenses require custom runtimes or source changes, and none can verify runtime integrity under an adversarial host. We present Walma, a framework for WebAssembly linear memory attestation that renders snapshots of linear memory as grayscale images and classifies them with a convolutional neural network. The network reads the row-aligned structure that compiled code imposes on memory and detects corruption that byte and texture statistics miss. On structured targets this extends to tampering that no program input triggers, such as direct memory writes by a malicious host. On real-world CVE-affected applications, hiding corruption from Walma's verdict requires overwriting a large fraction of memory, hundreds of kilobytes to megabytes on our targets. A Shannon-entropy analysis of benign memory bounds the class of out-of-band edits that the learned prior can detect. Across the 53 binaries of the WABench suite, attestation costs 1.07x–1.69x at the host boundary, making continuous, state-centric memory attestation for Wasm practical.