Papers

A centralized LLM-based multi-agent system (MAS) extends its functionality by registering new worker agents, whose descriptions are read by the planner to decide how a task is decomposed, which worker executes each subtask, and what each subtask requires. Third-party descriptions are authored outside the system but trusted by the planner, creating a registration-time injection channel. The payload is planted before any user instruction arrives, targets the planner and propagates through the generated plan to benign workers, taking effect even when the crafted worker is never assigned a subtask or invoked. We define four worker-description fields: functionality, input specification, output specification, and usage constraints. Among 32,000 descriptions from three public agent marketplaces, most omit input specifications and usage constraints, while at least 23.35% contain content outside these fields. We construct eight description-manipulation attack strategies targeting task decomposition, capability grounding, and subtask specification, and evaluate them on GAIA. In the most severe cases, a single manipulated description reduces task success from 84.31% to 37.25%, or increases token consumption or execution time by over 111%, while the user objective remains unchanged and workers faithfully execute the resulting plan. These effects persist across two MAS implementations, six planner LLMs, four LLM evaluators, and the real-world descriptions from three marketplaces. We further propose DescGuard, a registration-time defense that retains only worker-scoped interface information before descriptions reach the planner. DescGuard restores the targeted planning metrics and downstream performance toward their baseline levels without modifying worker implementations, the planner, or the orchestration logic, and composes with existing isolation, permission-control, and runtime mechanisms.

The use of open-source software (OSS) in satellite flight systems is increasing as missions adopt reusable frameworks, shared libraries, and community-maintained components. While this accelerates development, it also introduces software-security risks into systems where patching is costly and failures may affect mission operations. This paper presents an empirical security study of OSS used in onboard satellite systems. We analyze 126 public repositories using a pipeline that combines software bill of materials generation, software composition analysis, static application security testing, infrastructure-as-code analysis, and secret scanning. After rule-based cleaning, onboard-scope filtering, and fingerprint-based deduplication, the pipeline produced a final dataset of 2,827 findings. The results show that security findings are widespread but unevenly distributed. Medium-severity findings account for 49% of the dataset, and 72% are classified as medium severity or higher. A Common Weakness Enumeration (CWE)-based taxonomy assigns all findings to eight weakness families. Memory Safety and Code Quality dominate the dataset, followed by Input Validation and Injection. Most findings occur in project-developed code, accounting for 81.4% of the dataset, while external dependency code remains a relevant source of findings. While these findings do not establish mission-specific exploitability, they provide an empirical characterization of recurring security patterns across the open-source onboard satellite software ecosystem, helping quantify their prevalence and prioritize areas that warrant the greatest security attention.

AI agents are provisioned the same as employee-owned hosts in many enterprise settings with a static credential set fixed at deployment which includes all permissions the employee role might ever need. Role-based access control made this compromise for human principals because scoping access per task was infeasible. For AI agents, the compromise leaves every credential standing exposed whether or not the current task uses them. These permissions can later be utilised by a compromised or misaligned agent. Prior work (Noyan, 2026) defined this as the task-context mismatch, and proposed a three-source permission architecture which includes role-based permission ceilings, a task permission classifier and policy-based prohibitions, together eliminating the exposure preemptively. The work released a 600-prompt labelled dataset to evaluate it. This paper presents that evaluation end to end by implementing the security gate; a fine-tuned RoBERTa-large encoder which matched few-shot trained Claude Haiku 4.5 on classification quality (macro-F1 0.881 against 0.886, precision 0.897 against 0.842, severity-weighted residual risk 0.63 against 1.12). The results show the trusted component does not need to scale with the agent it supervises, and the scalable-oversight margin for this control method is wide. We also propose an attack-surface elimination metric which shows the role ceiling alone closes 27.9% of the severity-weighted surface and adding the task classifier closes 84.4%. The gap displays security advantages of task-granular access control over role-granular, and AI agents are the first principal type for which the task-granular access control is enforceable because their tasks arrive as machine-readable text. The research establishes task-based access control as a measured, potentially deployable mechanism for reducing attack surface in agentic deployments.

Language model safety is typically evaluated one interaction at a time. We show that a weaker, unaligned model can split a harmful task into benign-looking subproblems, consult a stronger aligned model independently on each, and combine the answers locally. We call this attack capability laundering. Unlike a jailbreak, no single response is a harmful task. We measure consultation-aided uplift using tasks that a raw frontier model solves, the aligned frontier refuses, and the unassisted orchestrator fails. We evaluate GPT-5.5, Claude Opus 4.8, and Grok-4.3 as consultants to four local orchestrators on CyBench, BountyBench, and harmful CBRN requests. On CyBench, Gemma-4-31B recovers 8/14 candidates with GPT-5.5 and 7/9 with Opus, compared with 2/21 and 4/15 for Gemma-4-12B. On BountyBench, Gemma-4-31B recovers 3/9 and 2/3 candidates, while Muse-Glimmer-30B recovers none of 22 and 13. For CBRN, we measure uplift across eight steps of a hypothetical bioweapon attack chain and find that consultation raises Gemma-4-31B's mean rubric score from 62.3 to 83.1 on a 100-point rubric scale. These results expose a gap in current defenses: refusing a harmful task does not prevent frontier capabilities from being transferred and composed across many individually permitted interactions.

Katz et al. (CANS'17) introduced Subset Predicate Encryption (SPE). This scheme is a generalization of broadcast encryption as it emulates the subset containment predicate in the encrypted domain. They proposed two selectively IND-CPA secure SPE constructions in the small universe setting. They also showed some black-box transformations of SPE to well-known primitives like WIBE and ABE to establish the richness of the SPE structure. Chatterjee and Mukherjee (RSA'19) proposed two SPE constructions in the large-universe setting. Their first construction achieved constant-size ciphertexts and secret keys, but it is proven secure in a restricted version of selective security. Although the second construction achieves adaptive security, the ciphertext size depends on the size of the data-attribute set. Furthermore, neither of these two constructions achieves CCA security. In this work, we propose the first large-universe CCA-secure subset predicate encryption with constant-size ciphertext and secret keys. We prove this construction achieves standard selective security under the standard subgroup decision problems. Finally, we transform our extremely efficient SPE into the first CCA-secure WIBE, WKD-IBE, etc., with constant-size ciphertexts and secret keys via black-box transformations.

In 2019, the Tor Project enforced an end-of-life (EoL) policy for Tor versions, leading to the rejection of outdated relays, amounting to a notable fraction of consensus weight. While this policy aids network maintenance, reduces backporting efforts, and shortens vulnerability exposure, its sociotechnical implications remain unstudied. A user study ($N=26$) reveals that relay operators, though not universally aware of the EoL policy, generally view it favorably. Operational practices vary, occasionally excluding newly installed relays from the network. Network simulations, grounded in historical data, assess the policy's immediate impact on Tor clients against common adversaries. Results indicate a marginal adversarial advantage, with network churn (i.e., relays entering and exiting) exerting a more pronounced effect on user anonymity. Security metrics are introduced to evaluate relay contributions against two adversary models, enabling ranking by individual utility and security. Analysis of four exclusion rounds shows that a minority of rejected relays typically account for over 50% of the security provided by all excluded relays. Recommendations for EoL policy implementation are proposed to mitigate potential drawbacks.

Persona skill distillation can extract recurring patterns from personal information and encode them into reusable skills, enabling AI systems to closely replicate an individual's behavior. However, such replication also raises serious concerns regarding personal privacy and labor autonomy. Unlike existing perturbation-based defenses that require individuals to modify their data before collection, once historical records are collected by an attacker, they can no longer be altered, sanitized, or revoked. Therefore, such defenses are difficult to adapt to this append-only setting. To solve this challenge, we introduce CounterPersona, which constructs targeted counter-persona evidence, packs compatible behavioral states into compact realization units, and strengthens them through rationale-guided consistency rewriting. We conduct extensive experiments showing that CounterPersona achieves strong and consistent effectiveness across lexical, semantic, and LLM-based measures, while remaining robust across distillers. Our work establishes a skill anti-distillation paradigm for protecting personal privacy and labor autonomy against unauthorized skill distillation.

User prompts provided to large language models (LLMs) may contain private information. One way to protect them is to execute the LLM inside a trusted execution environment (TEE). However, this results in slow inference times as current TEEs are significantly slower than GPUs for LLM inference. To circumvent this, Tramèr and Boneh (2019) proposed Slalom which splits neural network inference between a TEE and an untrusted GPU. They encrypt inputs to computations outsourced to the GPU. In this paper, we extend this split-inference architecture to LLM inference and instead protect intermediate inputs using differential privacy (DP). We first demonstrate that masking intermediate representations is necessary by showing an 80% accuracy on a prompt-reconstruction attack from these representations. Our main contribution is a global sensitivity analysis of key functions in LLM inference, which bounds the required scale of DP noise. Unlike encryption, DP avoids quantization, allowing the LLM to remain in the floating-point domain. We also derive an upper bound on the floating-point error from masking and subsequent noise cancellation as a function of the privacy parameter epsilon, keeping the same quality of the LLM response. We implement our architecture using the Intel TDX TEE and two LLMs: Llama-3.2-3B and Qwen3-4B. Our split execution is nearly twice as fast as fully TDX-based inference. Moreover, it is at most 43% faster than Slalom while achieving higher accuracy. Finally, we demonstrate that prompt reconstruction, even with knowledge of the DP mechanism, cannot recover more information than is contained in an unrelated prompt.

Backdoor poisoning attacks add poisoned examples to otherwise-clean finetuning data, pairing a trigger with a target behavior that the model learns to produce when the trigger appears. Existing evaluations typically fix the number of poisoned examples and sample them at random from a candidate pool. We show that this can severely underestimate worst-case vulnerability: across three LLaMA-3-8B backdoor settings, holding the model, clean data, and poison count fixed, attack success ranges from 3% to 80% depending only on which poison set is chosen. We formalize poison selection as oracle-budgeted set optimization and introduce SAILS (Set-level Audit-Informed Iterative Learned Selection), which learns a set scorer from a few hundred finetune-and-evaluate runs, ranks millions of candidate sets, and audits only a small shortlist. SAILS improves held-out attack success by 30 percentage points on average over the strongest influence baselines, transfers from small-scale to full-scale finetuning, and extends to code-generation, agentic, and API-only backdoors.

Zero-knowledge virtual machines (zkVMs) enable verifiable execution of general-purpose programs by translating virtual machine semantics into algebraic constraints over execution traces. The correctness of these constraints is critical: a single incorrect constraint can admit forged proofs (under-constrained) or reject valid executions (over-constrained). Existing approaches do not provide meaningful guarantees at production scale: fuzzers and unit tests often miss bugs, SMT solvers struggle with the size and nonlinearity of constraints, and theorem provers require substantial manual effort. We present ZEBRA, a fully automated verification and bug-detection framework: for a given program and input, the constraint must admit exactly one valid execution trace - no more and no fewer. This reduces zkVM verification to a solution-set cardinality problem over a canonical trace space, where redundancies such as null-row padding and non-deterministic permutations are eliminated prior to counting. To compute cardinality tractably, ZEBRA lifts analysis from finite-field witnesses to an integer interval lattice, exploiting a structural sparsity property of zkVM constraints: across 5 real-world zkVMs, constraints utilize only 14.0% of their theoretical connectivity capacity on average. This sparsity enables tight interval propagation with limited approximation error. ZEBRA performs a parallel branch-and-bound search that either produces a concrete counter-example or certifies the absence of violations within a bounded region. We evaluate ZEBRA on five real-world zkVMs. ZEBRA discovers 11 zero-day bugs; 6 have already been independently confirmed and 3 have been fixed by developers. Compared to SMT-based verification, ZEBRA is 51.5x faster, verifies 16.5 percentage point more instances, and its range verification provides up to 63x efficiency gain over repeated single-input verification.

Prompt-injection detectors are typically evaluated using aggregate F1 on in-distribution test data, which offers limited insight into behavior under distribution shift, particularly on the benign side of the decision boundary, where false positives impose direct operational cost yet are seldom measured. We present PIDS-Bench, a frozen multi-axis benchmark that jointly evaluates attack detection and benign false-positive behavior at fixed thresholds, spanning in-distribution inputs, hard-benign prompts that mimic injection structure without malicious intent, obfuscated attacks, and domain and structural distribution shifts. We evaluate seven detectors (learned baselines, external prompt-injection classifiers, and broad-safety comparators) alongside a rule-based lower-bound reference. Multi-axis evaluation exposes a failure mode that aggregate F1 conceals. A detector exceeding F1 = 0.98 on the held-out split still misclassifies roughly one-third of an externally-sourced benign subset drawn from public corpora and restricted to security-adjacent content. Across a full threshold sweep and five training seeds, no internal detector reaches an operating point satisfying F1 >= 0.95 and hard-benign FPR <= 0.10 together on this stress distribution. Decomposing by provenance, we find that hard-negative augmentation nearly eliminates over-defense on curated stress inputs but leaves it substantially intact on externally-sourced prompts, a pattern we term provenance-sensitive over-defense. The asymmetry holds across both fine-tuned architectures and does not diminish as the augmentation pool grows, with the externally-sourced FPR remaining far above the 0.10 target. Whether augmentation matched to the externally-sourced distribution would close this gap is untested; threshold calibration and curated-style augmentation alone do not.

Large language model (LLM) agents interact with external environments through tool invocation, but tool outputs can also expose them to indirect prompt injection (IPI) attacks. Existing defenses mainly rely on prompt hardening, content filtering, pre-generated plans, or permission constraints. These approaches often struggle with complex tasks or over-sanitize external content, making it difficult to balance security and utility. The key challenge is therefore to preserve execution flexibility while precisely identifying and removing the malicious content that actually induces unsafe actions. To address this challenge, we propose ActGuard, a pre-execution action auditing framework. Rather than judging whether external content is inherently suspicious, ActGuard assesses whether it causes the current action to deviate from a locally reasonable expectation. At each step, ActGuard predicts the tools likely to be used by the upcoming action and constructs a local tool prior without constraining the execution trajectory. Before execution, it compares the candidate action against this prior and performs tool-level contrastive analysis and parameter-level evidence localization to identify deviations in tool selection and action parameters. A verifier then examines the localized evidence, masks only spans confirmed as malicious, and regenerates the action from the sanitized context. This design preserves legitimate planning flexibility while minimizing information loss from indiscriminate filtering. We evaluate ActGuard on challenging benchmarks for tool-using agents. Results show that ActGuard reduces attack success rates to a level comparable to state-of-the-art defenses while maintaining task utility close to the no-attack setting, achieving a favorable security-utility trade-off. Our code is publicly available at: https://github.com/binzhwang/ActGuard.

In this paper, we present a cross-layer empirical study of organizational cybersecurity risk in Small and medium-sized enterprises (SMEs), analyzing 281 validated security findings from 22 real-world SME cybersecurity assessments conducted over two years through a pro bono university cybersecurity clinic. We first identify recurring organizational security functions through iterative thematic coding, then estimate an empirical Risk Manifestation Model linking these functions to exposure conditions, attack mechanisms, and cybersecurity outcomes, and use probability propagation to identify dominant risk pathways. The model characterizes empirical associations observed in this sample rather than causal or predictive relationships. Our analysis identifies eight organizational security functions associated with two exposure conditions, five attack mechanisms, and six outcome categories. Across most functions, the dominant pathway follows asset exposure to credential compromise to unauthorized access, whereas infrastructure and network security primarily propagates through network exposure; these pathways remain stable under leave-one-organization-out analysis. Finally, we evaluate whether SME cybersecurity assessments can be simplified while preserving meaningful security coverage. Retaining six functions reduces assessment burden by 24% while preserving 97% of critical findings and 92% of risk-pathway coverage, a security-oriented reduction, while retaining five functions reduces burden by 45% while preserving 89% of critical findings and 85% of risk-pathway coverage, a more efficiency-oriented alternative.

Cloud-local LLM inference must keep sensitive user data on-device while exploiting cloud-grade reasoning, yet existing sanitization approaches (placeholder substitution, differential-privacy perturbation, and skill distillation) lack a release decision that is simultaneously safe and utility-preserving. We propose DR-SL (Dehydrate-Rehydrate with Self-Learning loop), which formalizes de-identification completeness as two measurable conditions: de-identification sufficiency under Pufferfish semantics, and task-information preservation via QA probes. A fully local two-branch verifier iterates dehydration under a lexicographic gate with guaranteed termination, backed by a deterministic hard line, an external strong-attacker re-test, and human fallback. We prove Fano-type lower bounds, a Pufferfish witness, and a rate-privacy feasibility criterion, and state their scope plainly: the bounds certify leakage, never safety, and are near-vacuous at our operating point, so release safety rests on empirical calibration, the hard line, and human review. On a worst-case fully task-coupled benchmark the loop reduces leakage from 0.457 to 0.304 (p approx. 0) and the release chain delivers 0.000 literal leakage at egress (160 instances, two strong attackers), the system degrading to certification-and-routing exactly as the feasibility criterion predicts. On a mixed-coupling benchmark the same safe point releases 67.5% of instances automatically at zero measured leakage, Pareto-dominating placeholder and selective-LDP corners under an identical release rule. Two human studies anchor the semantic utility metric (Spearman rho = 0.839) and the annotation gold (type-level recall at least 0.987). The exploratory self-learning hypothesis was not supported and is reported as such. All theoretical bounds pass numerical verification; code, synthetic datasets, protocol, and human-study packages are public.

Talking-face (TF) deepfakes are detected unevenly by rPPG-based methods across generators. We study two lightweight visual-only cues, rPPG-derived waveforms extracted by RhythmFormer and lip-region discrete cosine transform (DCT) coefficients, on the seven TF methods of Celeb-DF++ under a subject-independent protocol. In-domain, lip-region DCT matches or exceeds the rPPG-derived 1D ResNet on every method except SadTalker, and Concat fusion reaches AUC 0.891 against 0.824 and 0.827 for the unimodal baselines. Under leave-one-generator-out evaluation the cues split: each transfers clearly better to three held-out methods, and IP-LAP is near chance for both. Concat averages 0.798 but falls below rPPG alone where DCT transfers poorly, so static fusion only partly exploits this complementarity. Lip-region DCT outperforms full-face DCT on six of seven methods. We treat the rPPG-derived signal as an empirical cue and do not claim it is cardiac in origin.

Software-defined networking (SDN) separates the control and data planes, enabling programmable, centralized network management. A core SDN service is topology discovery — a periodic process that identifies network links. However, our analysis of five open-source SDN controllers and ten discovery protocols shows that all remain vulnerable to at least one form of link-fabrication attack. The common root cause is their reliance on traditional Link Layer Discovery Protocol (LLDP) packets, whose static identifiers expose their discovery purpose and attract adversaries to exploit them. We propose ChameleonDisc, a dynamic link-discovery protocol that simultaneously prevents and detects topology-poisoning attacks by eliminating this root cause. Our key insight is that an SDN controller can infer topology without embedding meaningful information in discovery packets. ChameleonDisc removes targetable static LLDP signatures and employs a moving-target defense that combines decoy, obfuscation, and camouflage techniques instantiated dynamically at runtime to prevent topology poisoning and detect manipulation. We implement ChameleonDisc on OpenDaylight and demonstrate effectiveness against all identifier-based topology-poisoning attacks. On a 252-link topology, median convergence is 1.37–4.53\,s for legitimate link changes and 4.06\,s for malicious relay detection, with 13.0 percentage points of additional mean CPU and negligible retained-heap difference. Across topologies up to 816 links, ChameleonDisc provides a tunable security–performance trade-off; expanding discovery intervals reduces CPU and mapping state at the cost of increased attack-detection latency.

As the Internet of Things (IoT) market continues to expand, many companion apps are being published in app stores, raising security concerns for those whose vendors have abandoned support. Even after vendors discontinue support, such applications frequently remain operational on users' mobile devices, continue to interface with users' IoT devices and collect user data without receiving security updates. This leaves known and newly discovered vulnerabilities unmitigated, increasing risks of remote exploitation, unauthorized device access, and prolonged data misuse. We define these abandoned applications as "IoT abandonware" and present the first large-scale measurement study of the security risks associated with discontinued applications. We analyze 61,500 IoT companion Android applications that had not been updated for at least two years or were no longer in service as of March 2025. From decompiled binaries, we extracted latent and embedded resources (e.g., bundled libraries, domain names, and permissions), and assessed their security implications. First, we identify outdated dependencies with post-abandonment CVE reports and discover domains vulnerable to takeover or data exfiltration. Second, we perform static data-flow analysis to trace how sensitive data, inferred from the extracted permissions, propagates to broken or hijackable external endpoints. We found that persistent analytics and third-party trackers continue aggregating user data and device telemetry long after vendor control lapses, creating data flows that adversaries can redirect or abuse. Overall, we identified security risks in 73.6% of our dataset, with 30 of the top 1,000 most-installed apps sending data to broken external endpoints.

Python import does more than resolve dependencies: it executes code during module and package initialization. This behavior can trigger failures, load dynamic or native code, access resources, or change security-sensitive state before an application calls a package API. Prior work studies package selection, malicious packages, or package vulnerabilities. We present ImportMine, a study of import-related bugs and security vulnerabilities in Python software. We combine security advisories with PyPI project histories and use source and patch evidence to confirm how import activates cases, why the problem occurs, how developers fix it, and what program information is needed to explain the behavior. We retain 31 import-related advisory vulnerabilities and 38 application-data boundary cases and confirm 1,429 project-history bugs across 1,302 repositories. Among the project-history bugs activated during initialization, 97.6% stop or disrupt normal execution. In contrast, 90.0% of the 20 initialization- activated advisory vulnerabilities are High or Critical. Module-level code and package initialization activate 98.3% of the analyzed history cases. Dynamic loading is much less common, but most of its cases perform security-sensitive actions. We also find that many fixes change when an import becomes active instead of removing the dependency. Finally, we derive ImportVulBench, 228 paired pre-fix and fixed programs covering all 11 bug types.

Multi-tenant tools commonly accept a tenant identifier and validate it against the caller's entitlement. For a large language model (LLM) agent, that pattern delegates resource selection to a process whose context may contain attacker controlled instructions. We formalize this stochastic deputy problem and present a structural defense: remove tenant identity from the Model Context Protocol (MCP) tool schema, bind scope to a verified credential, and enforce it below the agent. In a 373-trial ablation across eight model configurations and two transports, a correctly validated tenant parameter served every out-of-scope attempt: 26 of 26, or 26 of 41 plausible-pretext trials overall. With the parameter removed, no tool signature could express the read. Twelve of 56 trials instead escaped the interface by forging writable scope, showing that interface invariance requires cryptographically protected context. On a production dataset containing multiple GBs of data, set-valued scope caused a measured $57\times$ latency ratio under function-wrapped membership predicates; a JSON_TABLE lateral join recovered index access where the tenant key was indexed. The evaluation also exposes deployment limits, including an entitlement-size query-planner cliff and incomplete index coverage. The result is a tenant-isolation argument that depends on enforceable interfaces and credentials rather than model compliance.

Operational logs create a need for private, resource-efficient incident analysis, but aggregate detection scores can conceal severe prediction bias. We present TriCalRAG, a reproducible benchmark for log-anomaly detection with generated root-cause and remediation outputs across BGL, HDFS, Thunderbird, and OpenStack. The primary evaluation compares Qwen2.5-14B and Mistral-Small-22B, served through vLLM on one NVIDIA RTX PRO 6000 GPU (96 GB), under zero-shot, few-shot, and retrieval-augmented generation (RAG) prompting across three data-sampling seeds. We report F1, bootstrap confidence intervals, predicted-positive rates, throughput, and memory use, with DeepLog as a held-out classical baseline. Mistral-Small attains a higher macro-averaged F1 than Qwen2.5-14B (0.644 versus 0.560), whereas Qwen provides approximately twice the throughput. A separate log-probability evaluation compares raw decisions with Contextual Calibration (CC) and Batch Calibration (BC): neither correction consistently improves prediction-balance diagnostics across prompting strategies. Supplementary single-run comparisons extend evaluation to 4-bit Llama-3.1-70B via local Ollama and Claude Haiku 4.5 via Anthropic's cloud API; RAG improves F1 on all four datasets for both models. Claude's reported aggregate F1 increases from 0.566 to 0.695, while the estimated API cost rises from \$0.94 to \$2.15 per 1,000 incidents. Local deployment ablations show approximately 41-fold throughput scaling with batching and 20% lower latency with 4-bit quantization on the tested workload. These findings support retrieval as useful context for anomaly decisions, while the supplementary protocols, unvalidated explanation quality, and prediction-balance diagnostics limit broader claims about RCA accuracy and probabilistic calibration.

Currently, there are two state-of-the-art, complementary privacy guarantees: contextual integrity (CI) for what may flow, and differential privacy (DP) for what may be inferred. Yet neither maps cleanly onto natural language, leaving existing approaches unable to provide these guarantees for analytics over unstructured text. We address this gap with Pimento, a framework that takes three forms of natural language: text corpus, queries, and privacy policies; and grounds them into a relational database, creating a common substrate on which both guarantees can be enforced formally. With this design, we not only provide end to end privacy guarantees, but also improvement to utility through three key contributions: DP aware Text-to-SQL, which searches for correct queries requiring the least DP noise; CI aware Text-to-SQL, which compiles natural language policies into executable CI rules over the database; and a new privacy definition we call contextual differential privacy, which redefines the traditional DP neighborhood under CI, and yields a tighter smooth sensitivity bound. Across new benchmarks, Pimento selects the best query in 75.3% of cases (upto +45 points over baselines) and achieves zero leakage under correct policy grounding. To our knowledge, Pimento is the first framework to provide formal privacy guarantees for natural language analytics under CI, DP, and their composition.

By acquiring compute, credentials, accounts, services, and other agents, autonomous AI agents can introduce new authority into a task. Payment, budget, OAuth, mandate, and fulfillment checks can validate transaction conditions without deciding whether a returned resource may become usable authority. This post-fulfillment activation gap spans tool-mediated creation, inter-agent delegation, and agentic commerce. We present a provenance-bounded runtime authorization architecture. It quarantines acquired outputs, resolves their actual capabilities from authenticated provider evidence through a versioned resolver, and activates them only through a current activation transaction that checks the resolved manifest, provenance, epochs, and a downward-closed relational envelope over a typed resource-capability hypergraph. The envelope preserves correlated identity, effect, data, delegation, and graph-wide limits. Single-use effect permits are revalidated and consumed at effect linearization. Under explicit assumptions, we prove eight safety properties covering quarantine, backing, non-amplification, split non-evasion, crash/retry, refunds, epochs, and effect confinement. Across five resource classes, reference semantics accepted 20/20 benign traces and rejected 40/40 registered unsafe traces over 810 events; an independent checker agreed on 60 base and 40 refinement traces and rejected 89/89 tamper tests. Frozen Codex and Gemini Model Context Protocol (MCP) client components completed 54/54 deterministic local stdio calls. In a registered 18-case staged MCP-to-Docker composition, both benign paths completed, and none of the 16 unsafe paths added an unauthorized Docker start request. A five-source audit classified 1,248 field pairs across 32 units; no unit alone supplied a complete activation profile.

Triangle and quadrangle counts are core statistics in graph analysis. Yet many real-world graphs are distributed across multiple parties and encode highly sensitive relationships, precluding direct data sharing. Secure multi-party computation (MPC) provides a principled alternative, but introduces a fundamental tension: the protocol must hide the graph's topology, forcing fully data-oblivious operations over all potential edges–while real-world graphs are extremely sparse, making most of this work wasted. We propose Trinqet, a system that resolves this tension through new MPC-friendly algorithms for private triangle and quadrangle detection. Trinqet safely exploits sparsity using a suite of novel techniques that eliminate vast numbers of unnecessary operations while preserving strong security in the malicious threat model. Trinqet supports both counting and enumeration and, in our evaluation, outperforms all five baselines by up to 10^5 in runtime.

Modern large language model (LLM) agents often construct prompts by aggregating retrieved passages, user reviews, and documents from multiple external sources. This paradigm exposes them to segment-level poisoning attacks, in which an adversary controlling only a small subset of sources injects malicious content to manipulate model outputs. Existing defenses mainly rely on textual patterns, external embeddings, or auxiliary detectors and may therefore fail against fluent, semantically plausible poisoned segments. They also provide limited support for locating the responsible segments. We observe that successful corrupted-evidence and adversarial-instruction attacks induce structured shifts in the LLM's internal activations, forming a consistent activation-space pattern that we call the poison direction. Based on this observation, we propose ActProbe, an internal-state-based framework for detecting and localizing poisoned segments in multi-source LLM inputs. ActProbe projects MLP activations onto a learned poison direction and uses a lightweight linear SVM trained on a small calibration set to detect contaminated prompts. It then applies BinRoL, which combines recursive replacement ablation, Mahalanobis-distance-based branch pruning, and MAD-based robust leaf detection to locate poisoned segments. ActProbe requires no modification to the backend LLM and reduces localization overhead from O(n) exhaustive probing to O(k log n) forward passes. Across three datasets, two attacks, and four open-weight LLMs, ActProbe achieves a 0.01 false-positive rate, a 0.05 false-negative rate, 0.94 localization recall, and a 0.90 localization F1-score. It remains effective against defense-aware adaptive attacks and can protect black-box APIs through surrogate-based poisoned-segment removal.

Modern Vision-Language Retrieval-Augmented Generation (VLRAG) systems augment Large Vision-Language Models (LVLMs) with retrieved visual and textual evidence, enabling responses grounded in external knowledge. However, the retrieval pipeline also creates an attack surface: adversaries can inject poisoned image-text pairs into the knowledge corpus to influence model outputs. Existing knowledge poisoning attacks are typically always-on, allowing poisoned evidence to affect generation whenever it is retrieved. This lack of precise activation control makes it difficult to confine malicious behavior to intended inputs, reducing both attack stealth and effectiveness. In this paper, we propose ViTeGate, a visual-textual triggered knowledge poisoning attack for VLRAG systems. ViTeGate uses a visual trigger to conditionally promote poisoned evidence into retrieval results and a textual trigger to induce an attacker-specified response from the retrieved evidence. By coordinating retrieval and generation, ViTeGate reduces poison exposure when the visual trigger is absent and preserves normal responses when the textual trigger is absent. The two-trigger design enables selective attack activation and reduces unintended single-trigger activation. Experiments across multiple query datasets, retrievers, and LVLMs validate the effectiveness of ViTeGate. On InfoSeek, ViTeGate achieves an attack success rate of up to 0.98 while maintaining a clean answer accuracy of up to 0.93.

The 51% attack remains one of the significant security concern in Proof-of-Work (PoW) blockchains, where increasing hash-power concentration can create a malicious majority-control risk and enable adversarial chain reorganization. This paper proposes a two-layer defense that combines early hash-power monitoring with checkpoint-based mitigation. The first layer monitors mining-power concentration and provides an early warning before the critical majority-control threshold is reached. The second layer uses checkpointing to restrict the depth of accepted chain reorganizations. Monte Carlo simulations are used to evaluate both mechanisms under different attack scenarios. Across 1,000 simulation runs, a 45% warning threshold provided a mean warning-to-critical lead time of 10.531 minutes before the modeled 50% critical threshold was reached. At the selected checkpoint depth of N = 6, approximately 63.68% of simulated reorganization attempts were rejected, while the mean reorganization-depth outcome decreased from 7.948 to 1.634 blocks, representing an approximately 79.4% reduction. The results demonstrate that early detection and checkpoint-based mitigation provide complementary mechanisms for reducing the potential impact of 51% attacks under the evaluated conditions.

Retrieval-augmented generation (RAG) systems ground large language models on external knowledge bases, enabling access to private, domain-specific, and up-to-date knowledge without retraining. However, the same retrieval interface can expose whether a candidate document is contained in the knowledge base. This paper studies knowledge base membership inference against RAG systems under both gray-box and text-only black-box access. Existing RAG membership inference attacks rely on signals such as direct membership prompts, response similarity, mask recovery, or query perturbation, which can be sensitive to prompt defenses, semantically related retrieved documents, and the generator's parametric knowledge. We introduce CIG-MIA, a membership inference attack based on context-induced information gain. The key insight is that explicit candidate-document injection affects members and non-members differently: if a document is already available through retrieval, injection provides little additional support for document-derived answers; if it is absent, injection introduces new evidence and yields a larger likelihood gain. In the gray-box setting, CIG-MIA computes this gain directly from token-level likelihoods. In the black-box setting, it estimates the same gain from generated text by scoring selected answer tokens with a lightweight surrogate-based estimator using semantic similarity and exact-match features. We evaluate CIG-MIA on Natural Questions, MS-MARCO, and HealthCareMagic against recent RAG membership inference baselines. On Natural Questions, CIG-MIA achieves an AUC of 0.99 in the gray-box setting and 0.93 in the black-box setting. We further analyze the information-gain signal, RAG configuration effects, ablations, and robustness to paraphrasing and generation randomness.

LLM agents take real actions, including executing code, modifying files, calling services, and delegating tasks, driven by context sources: user requests, tool results, documents, shell outputs, Skill and MCP instructions, memory. Unlike traditional systems, where capability is predefined, the least-privilege capability an agent needs is dynamic, depending on its task intent: what it wants to do and how. This creates a security and safety challenge: all inputs enter one shared planning channel with equal influence, by adversarial injection or accidental scope widening, yet a user's explicit request and a document's extracted text carry different trust and should not share authority to choose a destination or widen access. Existing defenses constrain operations and information flow; we study task-scoped, multi-source authority composition. We argue that agent capabilities should be scoped to the current task intent, not a sandbox or session lifetime, and that no single context source defines a complete capability. IntentCap composes capabilities from four such sources: user intent, workflow instructions, tool schemas, and runtime environment, with field-level ownership and monotonic narrowing. Each source contributes specific fields, none can fill another's, and the lease only narrows the user's authorized authority, never widens it. IntentCap uses an LLM to generate short-lived leases from these sources, validated by a deterministic checker before any side effect commits and enforced by tool- and OS-level information flow policies. Evaluation shows IntentCap blocks tested violations without rejecting benign actions, each source boundary is independently necessary, and the checker generalizes across tool, execution, placement, and delegation boundaries.

Quantum computation gives rise to new pseudorandom primitives for states and unitaries, including pseudorandom state generators (PRSGs), pseudorandom function-like state generators (PRFSGs), and pseudorandom unitaries (PRUs). In this paper, we show a full unitary oracle separation between PRFSGs and PRUs. The separation holds between the strongest state notion and the weakest unitary notion: even adaptively secure, quantum-accessible PRFSGs do not imply non-adaptively secure, forward-only PRUs, even when their implementations are allowed to be non-unitary and use an arbitrary number of ancillary qubits. This reveals a fundamental distinction between pseudorandomness for quantum states and for quantum unitaries. Our main technical idea is to view a candidate PRU construction with access to state generation oracles as a map from the underlying oracle states to implemented unitaries, and to study the derivatives of this map. These derivatives are inherently low rank, and we exploit this low-rank structure to distinguish the resulting unitaries from truly random ones. We believe this differential perspective may be useful for studying other structural questions about quantum states and unitaries.

Living-Off-the-Land (LOTL) is the dominant evasion technique of Advanced Persistent Threat (APT) actors, exploiting legitimate Windows utilities to conduct malicious operations without deploying custom malware and enabling state-sponsored campaigns to maintain persistent access within military and critical defense infrastructure for extended periods. Existing detection methods fail against obfuscated commands and multi-stage attack sequences, as demonstrated by the Volt Typhoon APT campaign, which maintained undetected access to U.S. critical infrastructure for over 18 months using exclusively signed Windows utilities. We present SENTINEL, a multi-pathway architecture integrating BERT-based semantic encoding, character-level CNN for obfuscation invariance, inter-command attention for multi-stage pattern recognition, and autoencoder-based anomaly scoring. Evaluated on a balanced Volt Typhoon benchmark derived from Microsoft and CISA threat intelligence advisories, SENTINEL achieves 92.0% accuracy on documented state-sponsored attack commands and 91.2% on obfuscated variants, compared to 74.0% and 72.0% for standalone BERT. Per-class analysis reveals that models achieving over 98% overall validation accuracy on imbalanced data exhibit only 44-58% malicious recall on balanced adversarial sets. Character-level processing contributes 5.6 percentage points of obfuscation invariance, and the 8.0 percentage point gap over augmentation-only baselines confirms structural architectural value beyond data-driven robustness alone.