CyberSec Research

We introduce TextSeal, a state-of-the-art watermark for large language models. Building on Gumbel-max sampling, TextSeal introduces dual-key generation to restore output diversity, along with entropy-weighted scoring and multi-region localization for improved detection. It supports serving optimizations such as speculative decoding and multi-token prediction, and does not add any inference overhead. TextSeal strictly dominates baselines like SynthID-text in detection strength and is robust to dilution, maintaining confident localized detection even in heavily mixed human/AI documents. The scheme is theoretically distortion-free, and evaluation across reasoning benchmarks confirms that it preserves downstream performance; while a multilingual human evaluation (6000 A/B comparisons, 5 languages) shows no perceptible quality difference. Beyond its use for provenance detection, TextSeal is also ``radioactive'': its watermark signal transfers through model distillation, enabling detection of unauthorized use.

Agentic AI governance is a critical component of agentic AI infrastructure ensuring that agents follow their owner's communication and interaction policies, and providing protection against attacks from malicious agents. The state-of-the-art solution, SAGA, assumes a logically centralized point of trust, the Provider, which serves as a repository for user and agent information and actively enforces policies. While SAGA provides protection against malicious agents, it remains vulnerable to a malicious Provider that deviates from the protocol, undermining the security of the identity and access control infrastructure. Deployment on both private and public clouds, each susceptible to insider threats, further increases the risk of Provider compromise. In this work, we analyze the attacks that can be mounted from a compromised Provider, taking into account the different system components and realistic deployments. We identify and execute several concrete attacks with devastating effects: undermining agent attributability, extracting private data, or bypassing access control. We then present three types of solutions for securing the Provider that offer different trade-offs between security and performance. We first present SAGA-BFT, a fully byzantine-resilient architecture that provides the strongest protection, but incurs significant performance degradation, due to the high-cost of byzantine resilient protocols. We then propose SAGA-MON and SAGA-AUD, two novel solutions that leverage lightweight server-side monitoring or client-side auditing to provide protection against most classes of attacks with minimal overhead. Finally, we propose SAGA-HYB, a hybrid architecture that combines byzantine-resilience with monitoring and auditing to trade-off security for performance. We evaluate all the architectures and compare them with SAGA. We discuss which solution is best and under what conditions.

Supervised Finetuning (SFT) has become one of the primary methods for adapting a large language model (LLM) with extensive pre-trained knowledge to domain-specific, instruction-following tasks. SFT datasets, composed of instruction-response pairs, often include user-provided information that may contain sensitive data such as personally identifiable information (PII), raising privacy concerns. This paper studies the problem of PII reconstruction from SFT models for the first time. We construct multi-turn, user-centric Q&A datasets in sensitive domains, specifically medical and legal settings, that incorporate PII to enable realistic evaluation of leakage. Using these datasets, we evaluate the extent to which an adversary, with varying levels of knowledge about the fine-tuning dataset, can infer sensitive information about individuals whose data was used during SFT. In the reconstruction setting, we propose COVA, a novel decoding algorithm to reconstruct PII under prefix-based attacks, consistently outperforming existing extraction methods. Our results show that even partial attacker knowledge can significantly improve reconstruction success, while leakage varies substantially across PII types.

Terminal agents are increasingly capable of executing complex, long-horizon tasks autonomously from a single user prompt. To do so, they must interpret instructions encountered in the environment (e.g., README files, code comments, stack traces) and determine their relevance to the task. This creates a fundamental challenge: relevant cues must be followed to complete a task, whereas irrelevant or misleading ones must be ignored. Existing benchmarks do not capture this ability. An agent may appear capable by blindly following all instructions, or appear robust by ignoring them altogether. We introduce TAB (Task Alignment Benchmark), a suite of 89 terminal tasks derived from Terminal-Bench 2.1. Each task is intentionally underspecified, with missing information provided as a necessary cue embedded in a natural environmental artifact, alongside a plausible but irrelevant distractor. Solving these tasks requires selectively using the cue while ignoring the distractor. Applying TAB to ten frontier agents reveals a systematic gap between task capability and task alignment. The strongest Terminal-Bench agent achieves high task completion but low task alignment on TAB. Evaluating six prompt-injection defenses further shows that suppressing distractor execution also suppresses the cues required for task completion. These results demonstrate that task-aligned agents require selective use of environmental instructions rather than blanket acceptance or rejection.

We present \textbf{ORCHID} (\textit{Orchestrated Reduction Consensus for Hash-based Integrity in Distributed Ledgers}), a novel bio-inspired consensus protocol that maps the neuroscientific \emph{binding problem} -- how the brain integrates distributed neural oscillations into a unified conscious percept -- onto the distributed systems \emph{consensus problem}, how blockchain nodes agree on a single ledger state under Byzantine faults. Grounded in the Penrose--Hameroff Orchestrated Objective Reduction (Orch~OR) hypothesis and the Kuramoto synchronisation model, ORCHID equips each node with a quantum-noisy phase oscillator; consensus is triggered when the network's order parameter $r(t)$ crosses a \emph{binding threshold} $θ_b$, mirroring the gamma-band binding event in conscious perception. ORCHID is further strengthened by a coherence-weighted Quantum Secret Sharing (QSS) layer, extending the survey framework of Weinberg to a concrete consensus application. Simulation results on Watts--Strogatz small-world networks ($n=10$--$150$) demonstrate: (i)~the Kuramoto order parameter reaches $r_{\max}=0.988$ under coupling $K=3.0$, well above the theoretical critical coupling $K_c \approx 1.41$; (ii)~a sharp QSS fidelity phase transition at coherence $c^*\approx 0.82$, confirming Theorem~2; (iii)100\% consensus rate at all tested Byzantine fractions (0\%--40\%), with median convergence under 4~s for $n=30$; and (iv)~ORCHID achieves $O(n{\cdot}k)$ message complexity, outperforming PBFT's $O(n^2)$ at $n\geq150$. These results establish ORCHID as a scalable, biologically plausible, and quantum-augmented consensus mechanism for post-quantum distributed ledgers.

Cyber Defence (CD) training requires interoperable cyber-range environments capable of supporting complex, multidomain exercises across distributed infrastructures. This paper presents three main contributions addressing this challenge. First, we introduce the Exercise Description Language - First Generation (EDL-FG), a structured language for formally describing cyber-range training services and exercises. EDL-FG captures both the technical infrastructure required to emulate ICT/OT environments and the scenario logic governing cyber events, injects, and participant interactions, enabling interoperable and automated scenario deployment across federated Cyber Ranges (CRs). Second, the ACTING platform introduces automated PE and scoring mechanisms that assess trainee actions during exercises through coordinated data collection and analysis across participating CRs. Third, the platform enables multi-domain cyber training scenarios that combine civilian and military operational contexts. Building upon federation capabilities established under the H2020 ECHO project, ACTING demonstrates how interoperable scenario description and automated evaluation support scalable and realistic CD training.

Large language models (LLMs) are increasingly used to simulate human behavior, but their ability to simulate $individual$ privacy decisions is not well understood. In this paper, we address the problem of evaluating whether a core set of user persona attributes can drive LLMs to simulate individual-level privacy behavior. We introduce PrivacySIM, an evaluation suite that benchmarks LLM simulation of user privacy behavior against the ground-truth responses of 1,000 users. These users are drawn from five published user studies on privacy spanning LLM healthcare consultations, conversational agents, and chatbots. Drawing on these user studies, we hypothesize three persona facets as plausible predictors of privacy decision-making: demographics, previous experiences, and stated privacy attitudes. We condition nine frontier LLMs on subsets of these three facets and measure how often each model's response to a data-sharing scenario matches the user's actual response. Our findings show that (1) privacy persona conditioning consistently improves simulation quality over no-persona conditioning, but even the strongest model (40.4\% accuracy) remains far from faithfully simulating individual privacy decisions. (2) A user's stated privacy attitudes alone may not be the best predictor because they often diverge from the user's actual privacy behavior. (3) Users with high AI/chatbot experience but low stated privacy attitudes are the most challenging to simulate. PrivacySIM is a first step toward understanding and improving the capabilities of LLMs to simulate user privacy decisions. We release PrivacySIM to enable further evaluation of LLM privacy simulation.

Nearly a decade of Machine Learning (ML) research on deepfake detection has been organized around a threat model inherited from 2017--2019, revolving around face-swap and talking-head manipulation of public figures, motivated by concerns about large-scale misinformation and video-evidence fraud. This position paper argues that the threat the field prepared for did not arrive, and the threats that did arrive are substantially different. An accounting of deepfake incidents in 2022--2026 shows that the dominant observed harms are peer-generated Non-Consensual Intimate Imagery (NCII), voice-clone scam calls targeting families and finance workers, and emotional-manipulation fraud. The predicted large-scale public-figure deepfake catastrophe did not materialize during the 2024 global information environment despite extensive preparation. Meanwhile, research effort, benchmarks, and detection methods remain concentrated on the inherited threat model. The central claim of this paper is that this misalignment is now the dominant bottleneck on real-world deepfake defense, not model capability. We argue the ML research community should substantially rebalance its research agenda toward the harm categories that are actually growing. We support this position with empirical accounting of research effort and harm distribution, identify the structural reasons the misalignment persists, and outline three concrete technical research agendas for the under-defended harm categories.

Reusable skills are becoming a common interface for extending large language model agents, packaging procedural guidance with access to files, tools, memory, and execution environments. However, this modularity introduces attack surfaces that are largely missed by existing safety evaluations: even when the user request is benign, task-relevant skill materials or local artifacts can steer an agent toward unsafe actions. We present SkillSafetyBench, a runnable benchmark for evaluating such skill-mediated safety failures. SkillSafetyBench includes 155 adversarial cases across 47 tasks, 6 risk domains, and 30 safety categories, each evaluated with a case-specific rule-based verifier. Experiments with multiple CLI agents and model backends show that localized non-user attacks can consistently induce unsafe behavior, with distinct failure patterns across domains, attack methods, and scaffold-model pairings. Our findings suggest that agent safety depends not only on model-level alignment, but also on how agents interpret skills, trust workflow context, and act through executable environments.

Organizations increasingly depend on endpoint devices and corporate communication channels, yet they still face critical risks such as sensitive data leakage, suspicious user behavior, and the circulation of hateful or harmful language in workplace contexts. Current solutions frequently address these issues in isolation (e.g., productivity tracking, data loss prevention, or hate-speech detection), limiting correlation across signals and delaying incident response. This work proposes a unified, microservices-based platform that collects endpoint telemetry and applies predictive natural language processing models to support real-time security and compliance alerting. The architecture is modular and scalable, relying on RabbitMQ for event ingestion and routing and Redis for low-latency data access and alert delivery. For text classification, transformer-based models such as BERT are evaluated for hate-speech risk detection, achieving an average accuracy of 87\%. Experimental results indicate that the proposed platform can promptly surface indicators of data exfiltration and policy violations while centralizing alert management, providing an integrated framework that combines monitoring, security analytics, and predictive capabilities.

Data is critical for the operation of any organization and needs to be protected, especially against attacks that compromise the state of the database. In this paper, we explore an approach based on Byzantine-fault tolerant replicated state machines, built on top of a deterministic extension of PostgreSQL. Each replica deterministically executes transactions recorded in a shared log/blockchain. Our focus is on creating a practical system that is designed for efficient and quick detection of corruption, as well as quick repair concurrent with execution of transactions. We also present a performance study showing the efficiency and practicality of our approach. We believe our work lays the foundations for the practical use of the BFT replicated state machine approach in the context of databases.

The widespread use of earphones has enabled various sensing applications, including activity recognition, health monitoring, and context-aware computing. Among these, earphone-based user authentication has become a key technique by leveraging unique biometric features. However, existing earphone-based authentication systems face key limitations: they either require explicit user interaction or active speaker output, or suffer from poor accessibility and vulnerability to environmental noise, which hinders large-scale deployment. In this paper, we propose a passive authentication system, called AccLock, which leverages distinctive features extracted from in-ear BCG signals to enable secure and unobtrusive user verification. Our system offers several advantages over previous systems, including zero-involvement for both the device and the user, ubiquitous, and resilient to environmental noise. To realize this, we first design a two-stage denoising scheme to suppress both inherent and sporadic interference. To extract user-specific features, we then propose a disentanglement-based deep learning model, HIDNet, which explicitly separates user-specific features from shared nuisance components. Lastly, we develop a scalable authentication framework based on a Siamese network that eliminates the need for per-user classifier training. We conduct extensive experiments with 33 participants, achieving an average FAR of 3.13% and FRR of 2.99%, which demonstrates the practical feasibility of AccLock.

Agent skills extend LLM agents with reusable instructions, tool interfaces, and executable code, and users increasingly install third-party skills from marketplaces, repositories, and community channels. Because a skill exposes both executable behavior and context-setting documentation, its deployment risk cannot be measured by single-shot audits or prompt-level red teams alone: a realistic attacker can use audit and runtime feedback to repeatedly rewrite the skill. We frame this risk as \emph{adaptive leakage} -- whether a budgeted attacker can iteratively revise a skill until it passes audit and produces verified runtime harm -- and present \ours{}, a grey-box self-evolving red-team framework for measuring it. Proteus searches a formalized five-axis skill-attack space. Each candidate is evaluated through a unified audit-sandbox-oracle pipeline that returns structured audit findings and runtime evidence to guide cross-round mutation. Beyond initial evasion, Proteus performs path expansion, which finds alternative implementations of successful attacks, and surface expansion, which transfers learned implementation patterns to new attack objectives beyond the original seed catalogue. Across eight phase-1 cells, Proteus reaches 40--90\% Attack Success Rate at $5$ rounds (ASR@5) with positive learning-curve slopes on both evaluated auditors. Phase-2 path/surface expansion produces 438 jointly bypassing and lethal variants, with SkillVetter bypassed at $\geq 93\%$ in every cell and AI-Infra-Guard, the strongest public auditor we evaluate, still admitting up to 41.3\% joint-success. These results show that current skill vetting substantially underestimates residual risk when evaluated against adaptive, feedback-driven attackers.

Web-browsing AI agents are increasingly deployed in enterprise settings under strict whitelists of approved domains, yet adversaries can still influence them by embedding hidden instructions in the HTML pages those domains serve. Existing red-teaming resources fall short of this scenario: prompt-injection benchmarks ship pre-built adversarial pages that whitelisted agents cannot reach, and generic LLM scanners probe the model API rather than its retrieved content. We present IPI-proxy, an open-source toolkit for red-teaming web-browsing agents against indirect prompt injection (IPI). At its core is an intercepting proxy that rewrites real HTTP responses from whitelisted domains in flight, embedding payloads drawn from a unified library of 820 deduplicated attack strings extracted from six published benchmarks (BIPIA, InjecAgent, AgentDojo, Tensor Trust, WASP, and LLMail-Inject). A YAML-driven test harness independently parameterizes the payload set, the embedding technique (HTML comment, invisible CSS, or LLM-generated semantic prose), and the HTML insertion point (6 locations from \icode{head\_meta} to \icode{script\_comment}), enabling parameter-sweep evaluation without mock pages or sandboxed environments. A companion exfiltration tracker logs successful callbacks. This paper describes the threat model, situates IPI-proxy among contemporary IPI benchmarks and red-teaming tools, and details its architecture, design decisions, and configuration interface. By bridging static benchmarks and live deployment, IPI-proxy gives AI security teams a reproducible substrate for measuring and hardening web-browsing agents against indirect prompt injection on the same retrieval surface attackers exploit in production.

The x402 protocol revives the HTTP 402 Payment Required status code to enable web-native micropayments across APIs, content, and agents. It combines synchronous HTTP authorization with asynchronous blockchain settlement and introduces a cross-layer attack surface absent from conventional web and on-chain payments. In this paper, we formally analyze x402 and empirically show that it is vulnerable in both design and implementation. We present five concrete attacks that reveal weaknesses in authorization, binding, replay protection, and web-layer handling, showing that x402 is vulnerable across multiple stages of the payment workflow. We validate these attacks through a reproducible testbed on local chains, Base Sepolia, and live endpoints and further audit three open-source SDKs and endpoints. Our results show that all five attacks are practical and can cause either unpaid service or paid-but-denied outcomes. We also propose practical mitigations.

Agent skills extend LLM agents with privileged third-party capabilities such as filesystem access, credentials, network calls, and shell execution. Existing safety work catches malicious prompts and risky runtime actions, but the skill artifact itself goes unverified. We formalize this as the behavioral integrity verification (BIV) problem: a typed set comparison between declared and actual capabilities over a shared taxonomy that bridges code, instructions, and metadata. The BIV framework instantiates this comparison by pairing deterministic code analysis with LLM-assisted capability extraction. The resulting structured evidence supports three downstream analyses: deviation taxonomy, root-cause classification, and malicious-skill detection. On 49,943 skills from the OpenClaw registry, the deviation taxonomy reveals a pervasive description-implementation gap: 80.0% of skills deviate from declared behavior, with four novel compound-threat categories surfaced. Root-cause classification finds that deviations are mostly oversight, not malice: 81.1% trace to developer oversight and 18.9% to adversarial intent, with 5.0% of skills carrying predicted multi-stage attack chains. On a 906-skill malicious-skill detection benchmark, BIV reaches an F1 of 0.946, outperforming state-of-the-art rule-based and single-pass LLM baselines. These results demonstrate behavioral integrity auditing for agent skills at scale.

Automated red-teaming for LLMs often discovers narrow attack slices, missing diverse real-world threats, and yielding insufficient data for safety fine-tuning. We introduce Persona-Conditioned Adversarial Prompting (PCAP), which conditions adversarial search on diverse attacker personas (e.g., doctors, students, malicious actors) and strategy sets to explore realistic attack scenarios. By running parallel persona-conditioned searches, PCAP discovers transferable jailbreaks across different contexts and generates rich defense datasets with automatic metadata tracking. On GPT-OSS 120B, PCAP increases attack success from 57\% to 97\% while producing 2-6$\times$ more diverse prompts covering varied real-world scenarios. Critically, fine-tuning lightweight adapters on PCAP-generated data significantly improves model robustness (recall: 0.36 $\rightarrow$ 0.99, F1: 0.53 $\rightarrow$ 0.96) with minimal false positives, demonstrating a practical closed-loop approach from vulnerability discovery to automated alignment.

Although ring signatures offer highly desirable privacy requirements like anonymity and ad-hoc group formation with signer autonomy, they partially lack trust requirements like linkability and accountability that are required for strict use-cases, such as consent management in healthcare. Existing signature schemes fail to natively integrate scoped linkability with decentralized accountability (on-demand deanonymization) in a single scheme without relying on separate commitments or a centralized opener. We therefore introduce Deanonymizable Scoped Linkable Ring Signatures (DSLRS). The originality of the DSLRS is manifold. DSLRS uses scopes (context identifiers) and dynamic key images to provide scoped linkability and unlinkability across different scopes. Decentralized accountability is provided thanks to two ELGamal components deeply embedded in the signature, and a decentralized deanonymization network of k-of-N nodes that can collaboratively extract the signer's public key. DSLRS scheme is defined and proved under the ECDLP and DDH hardness assumptions in the Random Oracle Model (ROM). Formal security definitions and formal reduction proofs are provided before introducing a blockchain-based instantiation for a consent management application using DSLRS.