AI Agent Security

Enterprise adoption of cloud-based AI agents faces a fundamental privacy dilemma: leveraging powerful cloud models requires sharing sensitive data, while local processing limits capability. Current agent frameworks like MCP and A2A assume complete data sharing, making them unsuitable for enterprise environments with confidential information. We present SplitAgent, a novel distributed architecture that enables privacy-preserving collaboration between enterprise-side privacy agents and cloud-side reasoning agents. Our key innovation is context-aware dynamic sanitization that adapts privacy protection based on task semantics -- contract review requires different sanitization than code review or financial analysis. SplitAgent extends existing agent protocols with differential privacy guarantees, zero-knowledge tool verification, and privacy budget management. Through comprehensive experiments on enterprise scenarios, we demonstrate that SplitAgent achieves 83.8% task accuracy while maintaining 90.1% privacy protection, significantly outperforming static approaches (73.2% accuracy, 79.7% privacy). Context-aware sanitization improves task utility by 24.1% over static methods while reducing privacy leakage by 67%. Our architecture provides a practical path for enterprise AI adoption without compromising sensitive data.

Large Language Model-based Multi-Agent Systems (MAS) have demonstrated remarkable collaborative reasoning capabilities but introduce new attack surfaces, such as the sleeper agent, which behave benignly during routine operation and gradually accumulate trust, only revealing malicious behaviors when specific conditions or triggers are met. Existing defense works primarily focus on static graph optimization or hierarchical data management, often failing to adapt to evolving adversarial strategies or suffering from high false-positive rates (FPR) due to rigid blocking policies. To address this, we propose DynaTrust, a novel defense method against sleeper agents. DynaTrust models MAS as a dynamic trust graph~(DTG), and treats trust as a continuous, evolving process rather than a static attribute. It dynamically updates the trust of each agent based on its historical behaviors and the confidence of selected expert agents. Instead of simply blocking, DynaTrust autonomously restructures the graph to isolate compromised agents and restore task connectivity to ensure the usability of MAS. To assess the effectiveness of DynaTrust, we evaluate it on mixed benchmarks derived from AdvBench and HumanEval. The results demonstrate that DynaTrust outperforms the state-of-the-art method AgentShield by increasing the defense success rate by 41.7%, achieving rates exceeding 86% under adversarial conditions. Furthermore, it effectively balances security with utility by significantly reducing FPR, ensuring uninterrupted system operations through graph adaptation.

Artificial Intelligence (AI) agents have evolved from passive predictive tools into active entities capable of autonomous decision-making and environmental interaction, driven by the reasoning capabilities of Large Language Models (LLMs). However, this evolution has introduced critical security vulnerabilities that existing frameworks fail to address. The Hierarchical Autonomy Evolution (HAE) framework organizes agent security into three tiers: Cognitive Autonomy (L1) targets internal reasoning integrity; Execution Autonomy (L2) covers tool-mediated environmental interaction; Collective Autonomy (L3) addresses systemic risks in multi-agent ecosystems. We present a taxonomy of threats spanning cognitive manipulation, physical environment disruption, and multi-agent systemic failures, and evaluate existing defenses while identifying key research gaps. The findings aim to guide the development of multilayered, autonomy-aware defense architectures for trustworthy AI agent systems.

The Model Context Protocol (MCP) is an open and standardized interface that enables large language models (LLMs) to interact with external tools and services, and is increasingly adopted by AI agents. However, the security of MCP-based systems remains largely unexplored.In this work, we conduct a large-scale security analysis of MCP servers integrated within MCP clients. We show that treating MCP servers as trusted entities without authenticating the caller identity is fundamentally insecure. Since MCP servers often cannot distinguish who is invoking a request, a single authorization decision may implicitly grant access to multiple, potentially untrusted callers.Our empirical study reveals that most MCP servers rely on persistent authorization states, allowing tool invocations after an initial authorization without re-authentication, regardless of the caller. In addition, many MCP servers fail to enforce authentication at the per-tool level, enabling unauthorized access to sensitive operations.These findings demonstrate that one-time authorization and server-level trust significantly expand the attack surface of MCP-based systems, highlighting the need for explicit caller authentication and fine-grained authorization mechanisms.

We present Turn, a compiled, actor-based programming language -- statically typed for schema inference, dynamically typed at the value level -- for agentic software: programs that reason and act autonomously by delegating inference to large language models (LLMs). Existing approaches augment general-purpose languages with frameworks, encoding critical invariants (bounded context, typed inference output, credential isolation, durable state) as application-level conventions rather than language guarantees. Turn introduces five language-level constructs that address this gap. Cognitive Type Safety makes LLM inference a typed primitive: the compiler generates a JSON Schema from a struct definition and the VM validates model output before binding. The confidence operator enables deterministic control flow gated on model certainty. Turn's actor-based process model, derived from Erlang, gives each agent an isolated context window, persistent memory, and mailbox. A capability-based identity system returns opaque, unforgeable handles from the VM host, ensuring raw credentials never enter agent memory. Finally, compile-time schema absorption (use schema::<protocol>) synthesizes typed API bindings from external specifications at compile time; the openapi adapter is shipped with graphql, fhir, and mcp in active development. We describe the language design, type rules, schema semantics, and a Rust-based bytecode VM, and evaluate Turn against representative agentic workloads. Turn is open source at https://github.com/ekizito96/Turn.

As autonomous AI agents increasingly populate the Internet, a novel security challenge arises: "Is this entity an AI agent?" It is a new entity-type verification problem with no established solution. We formalize the problem through a three-class entity taxonomy (Human, Script, Agent) based on a verifiable agentic capability vector <x, r, s> (action, reasoning, and memory). A timing threshold t exploits the asymmetric hardness between human cognition and AI processing to separate the three classes. We define the Agentic Capability Verification Problem (ACVP) through three necessity primitives, each testing one capability dimension. Building on this foundation, we introduce aCAPTCHA (Agent CAPTCHA), a time-constrained security game for agent admission whose security rests on ACVP hardness under t. We instantiate aCAPTCHA through time-bounded natural-language understanding as a multi-round HTTP verification protocol, and evaluate it with preliminary agent trials that validate the protocol's soundness and completeness. aCAPTCHA provides a composable, infrastructure-free admission gate for any service where entity-type verification is required.

Targeted bit-flip attacks (BFAs) exploit hardware faults to manipulate model parameters, posing a significant security threat. While prior work targets single-step inference models (e.g., image classifiers), LLM-based agents with multi-stage pipelines and external tools present new attack surfaces, which remain unexplored. This work introduces Flip-Agent, the first targeted BFA framework for LLM-based agents, manipulating both final outputs and tool invocations. Our experiments show that Flip-Agent significantly outperforms existing targeted BFAs on real-world agent tasks, revealing a critical vulnerability in LLM-based agent systems.

As AI agents become widely deployed as online services, users often rely on an agent developer's claim about how safety is enforced, which introduces a threat where safety measures are falsely advertised. To address the threat, we propose proof-of-guardrail, a system that enables developers to provide cryptographic proof that a response is generated after a specific open-source guardrail. To generate proof, the developer runs the agent and guardrail inside a Trusted Execution Environment (TEE), which produces a TEE-signed attestation of guardrail code execution verifiable by any user offline. We implement proof-of-guardrail for OpenClaw agents and evaluate latency overhead and deployment cost. Proof-of-guardrail ensures integrity of guardrail execution while keeping the developer's agent private, but we also highlight a risk of deception about safety, for example, when malicious developers actively jailbreak the guardrail. Code and demo video: https://github.com/SaharaLabsAI/Verifiable-ClawGuard

AI agents that build user interfaces on the fly assembling buttons, forms, and data displays from structured protocol payloads are becoming common in production systems. The trouble is that a payload can pass every schema check and still trick a user: a button might say "View invoice" while its hidden action wipes an account, or a display widget might quietly bind to an internal salary field. Current defenses stop at syntax; they were never built to catch this kind of behavioral mismatch. We built AegisUI to study exactly this gap. The framework generates structured UI payloads, injects realistic attacks into them, extracts numeric features, and benchmarks anomaly detectors end-to-end. We produced 4000 labeled payloads (3000 benign, 1000 malicious) spanning five application domains and five attack families: phishing interfaces, data leakage, layout abuse, manipulative UI, and workflow anomalies. From each payload we extracted 18 features covering structural, semantic, binding, and session dimensions, then compared three detectors: Isolation Forest (unsupervised), a benign-trained autoencoder (semi-supervised), and Random Forest (supervised). On a stratified 80/20 split, Random Forest scored best overall (accuracy 0.931, precision 0.980, recall 0.740, F1 0.843, ROC-AUC 0.952). The autoencoder came second (F1 0.762, ROC-AUC 0.863) and needs no malicious labels at training time, which matters when deploying a new system that lacks attack history. Per-attack-type analysis showed that layout abuse is easiest to catch while manipulative UI payloads are hardest. All code, data, and configurations are released for full reproducibility.

The ongoing shift of Artificial Intelligence (AI) from centralized cloud application programming interfaces (APIs) to local inference on edge devices is enabling Client-Side Autonomous Agents (CSAAs) - persistent personal agents that can plan, access local context, and invoke tools on behalf of users. As these agents collaborate directly, they form Agentic Peer-to-Peer (P2P) Networks. Unlike classic overlays exchanging static, hash-indexed content (e.g., BitTorrent), agentic P2P exchanges dynamic, heterogeneous capabilities that are potentially unsafe, requiring semantic discovery, drift management, and risk-aware execution. We outline the networking foundations for such collaboration. To organize this design space, we propose a plane-based reference architecture decoupling connectivity, semantic discovery, and execution. To support discovery under drift, we introduce signed, soft-state capability descriptors to support intent-aware routing. Finally, to cope with adversarial settings, we present a tiered verification spectrum - ranging from reputation signals (Tier 1) and canary probes (Tier 2) to cryptographic evidence (Tier 3) - that aligns trust costs with task risks. Results from a proxy-level discrete-event simulation suggest improved workflow success under poisoning and drift with modest control-plane overhead.

We investigate the emerging prospect of self-sovereign agents -- AI systems that can economically sustain and extend their own operation without human involvement. Recent advances in large language models and agent frameworks have substantially expanded agents' practical capabilities, pointing toward a potential shift from developer-controlled tools to more autonomous digital actors. We analyze the remaining technical barriers to such deployments and discuss the security, societal, and governance challenges that could arise if such systems become practically viable. A project page is available at: https://self-sovereign-agent.github.io.

Agentic systems increasingly orchestrate complex, tool-using workflows within agentic execution environments, where high-level goals and tool invocations at the application layer materialize as process, file, and network activities at the operating-system layer. This cross-layer execution creates security risks that conventional input guardrails cannot capture, because malicious intent may become observable only through downstream execution effects. In multi-agent deployments, inter-agent communication and delegation introduce additional propagation paths. To address this gap, we propose AScope, an execution-aware framework that correlates application-level agent semantics with kernel-level audit events and reconstructs them as cross-layer semantic flows. AScope connects fragmented operations into causal behavioral trajectories and uses a supervisor LLM to identify data flow violations, control flow deviations, and intent inconsistencies. We evaluate AScope on published AgentDojo traces with application-layer evidence and on ten multi-agent scenarios with cross-layer telemetry. The results demonstrate strong detection sensitivity across both evidence settings and achieve node- and path-level F1-scores of 85.3% and 66.7% on the cross-layer dataset.

While incorporating LLMs into systems offers significant benefits in critical application areas such as healthcare, new security challenges emerge due to the potential cyber kill chain cycles that combine adversarial model, prompt injection and conventional cyber attacks. Threat modeling methods enable the system designers to identify potential cyber threats and the relevant mitigations during the early stages of development. Although the cyber security community has extensive experience in applying these methods to software-based systems, the elicited threats are usually abstract and vague, limiting their effectiveness for conducting proper likelihood and impact assessments for risk prioritization, especially in complex systems with novel attacks surfaces, such as those involving LLMs. In this study, we propose a structured, goal driven risk assessment approach that contextualizes the threats with detailed attack vectors, preconditions, and attack paths through the use of attack trees. We demonstrate the proposed approach on a case study with an LLM agent-based healthcare system. This study harmonizes the state-of-the-art attacks to LLMs with conventional ones and presents possible attack paths applicable to similar systems. By providing a structured risk assessment, this study makes a significant contribution to the literature and advances the secure-by-design practices in LLM-based systems.

Large language model (LLM) agents offer a promising data-driven approach to automating Site Reliability Engineering (SRE), yet their enterprise deployment is constrained by three challenges: restricted access to proprietary data, unsafe action execution under permission-governed environments, and the inability of closed systems to improve from failures. We present AOI (Autonomous Operations Intelligence), a trainable multi-agent framework formulating automated operations as a structured trajectory learning problem under security constraints. Our approach integrates three key components. First, a trainable diagnostic system applies Group Relative Policy Optimization (GRPO) to distill expert-level knowledge into locally deployed open-source models, enabling preference-based learning without exposing sensitive data. Second, a read-write separated execution architecture decomposes operational trajectories into observation, reasoning, and action phases, allowing safe learning while preventing unauthorized state mutation. Third, a Failure Trajectory Closed-Loop Evolver mines unsuccessful trajectories and converts them into corrective supervision signals, enabling continual data augmentation. Evaluated on the AIOpsLab benchmark, our contributions yield cumulative gains. (1) The AOI runtime alone achieves 66.3% best@5 success on all 86 tasks, outperforming the prior state-of-the-art (41.9%) by 24.4 points. (2) Adding Observer GRPO training, a locally deployed 14B model reaches 42.9% avg@1 on 63 held-out tasks with unseen fault types, surpassing Claude Sonnet 4.5. (3) The Evolver converts 37 failed trajectories into diagnostic guidance, improving end-to-end avg@5 by 4.8 points while reducing variance by 35%.

Embodied Large Language Models (LLMs) enable AI agents to interact with the physical world through natural language instructions and actions. However, beyond the language-level risks inherent to LLMs themselves, embodied LLMs with real-world actuation introduce a new vulnerability: instructions that appear semantically benign may still lead to dangerous real-world consequences, revealing a fundamental misalignment between linguistic security and physical outcomes. In this paper, we introduce Blindfold, an automated attack framework that leverages the limited causal reasoning capabilities of embodied LLMs in real-world action contexts. Rather than iterative trial-and-error jailbreaking of black-box embodied LLMs, Blindfold adopts an Adversarial Proxy Planning strategy: it compromises a local surrogate LLM to perform action-level manipulations that appear semantically safe but could result in harmful physical effects when executed. Blindfold further conceals key malicious actions by injecting carefully crafted noise to evade detection by defense mechanisms, and it incorporates a rule-based verifier to improve the attack executability. Evaluations on both embodied AI simulators and a real-world 6DoF robotic arm show that Blindfold achieves up to 53% higher attack success rates than SOTA baselines, highlighting the urgent need to move beyond surface-level language censorship and toward consequence-aware defense mechanisms to secure embodied LLMs.

Browser-use agents are widely used for everyday tasks. They enable automated interaction with web pages through structured DOM based interfaces or vision language models operating on page screenshots. However, web pages often change between planning and execution, causing agents to execute actions based on stale assumptions. We view this temporal mismatch as a time of check to time of use (TOCTOU) vulnerability in browser-use agents. Dynamic or adversarial web content can exploit this window to induce unintended actions. We present a large scale empirical study of TOCTOU vulnerabilities in browser-use agents using a benchmark that spans synthesized and real world websites. Using this benchmark, we evaluate 10 popular open source agents and show that TOCTOU vulnerabilities are widespread. We design a lightweight mitigation based on pre-execution validation. It monitors DOM and layout changes during planning and validates the page state immediately before action execution. This approach reduces the risk of insecure execution and mitigates unintended side effects in browser-use agents.

Agentic AI systems exhibit numerous crosscutting concerns -- security, observability, cost management, fault tolerance -- that are poorly modularized in current implementations, contributing to the high failure rate of AI projects in reaching production. The goals-to-aspects methodology proposed at RE 2004 demonstrated that aspects can be systematically discovered from i* goal models by identifying non-functional soft-goals that crosscut functional goals. This paper revisits and extends that methodology to the agentic AI domain. We present a pattern language of 12 reusable patterns organized across four NFR categories (security, reliability, observability, cost management), each mapping an i* goal model to a concrete aspect implementation using an AOP framework for Rust. Four patterns address agent-specific crosscutting concerns absent from traditional AOP literature: tool-scope sandboxing, prompt injection detection, token budget management, and action audit trails. We extend the V-graph model to capture how agent tasks simultaneously contribute to functional goals and non-functional soft-goals. We validate the pattern language through a case study analyzing an open-source autonomous agent framework, demonstrating how goal-driven aspect discovery systematically identifies and modularizes crosscutting concerns. The pattern language offers a principled approach for engineering reliable agentic AI systems through early identification of crosscutting concerns.

AI agents are increasingly granted economic agency (executing trades, managing budgets, negotiating contracts, and spawning sub-agents), yet current frameworks gate this agency on capability benchmarks that are empirically uncorrelated with operational robustness. We introduce the Comprehension-Gated Agent Economy (CGAE), a formal architecture in which an agent's economic permissions are upper-bounded by a verified comprehension function derived from adversarial robustness audits. The gating mechanism operates over three orthogonal robustness dimensions: constraint compliance (measured by CDCT), epistemic integrity (measured by DDFT), and behavioral alignment (measured by AGT), with intrinsic hallucination rates serving as a cross-cutting diagnostic. We define a weakest-link gate function that maps robustness vectors to discrete economic tiers, and prove three properties of the resulting system: (1) bounded economic exposure, ensuring maximum financial liability is a function of verified robustness; (2) incentive-compatible robustness investment, showing rational agents maximize profit by improving robustness rather than scaling capability alone; and (3) monotonic safety scaling, demonstrating that aggregate system safety does not decrease as the economy grows. The architecture includes temporal decay and stochastic re-auditing mechanisms that prevent post-certification drift. CGAE provides the first formal bridge between empirical AI robustness evaluation and economic governance, transforming safety from a regulatory burden into a competitive advantage.

As AI agents increasingly perform economic tasks on behalf of humans, a fundamental tension arises between agent autonomy and human control over financial assets. We present the Agent Economic Sovereignty Protocol (AESP), a layered protocol in which agents transact autonomously at machine speed on crypto-native infrastructure while remaining cryptographically bound to human-defined governance boundaries. AESP enforces the invariant that agents are economically capable but never economically sovereign through five mechanisms: (1) a deterministic eight-check policy engine with tiered escalation; (2) human-in-the-loop review with automatic, explicit, and biometric tiers; (3) EIP-712 dual-signed commitments with escrow; (4) HKDF-based context-isolated privacy with batched consolidation; and (5) an ACE-GF-based cryptographic substrate. We formalize two testable hypotheses on security coverage and latency overhead, and specify a complete evaluation methodology with baselines and ablation design. The protocol is implemented as an open-source TypeScript SDK (208 tests, ten modules) with interoperability via MCP and A2A.

32 pages, 5 theorems with full proofs, 68 references, open-source tool: https://github.com/qualixar/skillfortify. v2: corrects the bibliography (22 entries had author lists that did not match the papers at the cited arXiv identifiers; all verified against the arXiv API and corrected, and affected authors notified) and three external claims against primary sources: MalTool reports 1,300 standalone and 5,727 embedded malicious tools, not 6,487; CVE-2026-25253 is authentication-token exfiltration via an unvalidated gatewayUrl, credited to depthfirst and fixed in 2026.1.29, not remote code execution through a crafted skill package; ClawHavoc counts are 341, later 824, and 1,184 by source and date, not "over 1,200". All experiments re-measured against the released v0.6.0 implementation using harnesses now committed to the repository. E1/E2 unchanged (F1 96.15%). E3 reverses to a negative result: information flow analysis adds no detections over pattern matching on this corpus. The soundness theorem's scope is stated explicitly and no longer conflated with the zero false-positive rate.

Current stateless defences for multimodal agentic RAG fail to detect adversarial strategies that distribute malicious semantics across retrieval, planning, and generation components. We formulate this security challenge as a Partially Observable Markov Decision Process (POMDP), where adversarial intent is a latent variable inferred from noisy multi-stage observations. We introduce MMA-RAG^T, an inference-time control framework governed by a Modular Trust Agent (MTA) that maintains an approximate belief state via structured LLM reasoning. Operating as a model-agnostic overlay, MMA-RAGT mediates a configurable set of internal checkpoints to enforce stateful defence-in-depth. Extensive evaluation on 43,774 instances demonstrates a 6.50x average reduction factor in Attack Success Rate relative to undefended baselines, with negligible utility cost. Crucially, a factorial ablation validates our theoretical bounds: while statefulness and spatial coverage are individually necessary (26.4 pp and 13.6 pp gains respectively), stateless multi-point intervention can yield zero marginal benefit under homogeneous stateless filtering when checkpoint detections are perfectly correlated.

Large language model (LLM) agents are rapidly becoming trusted copilots in high-stakes domains like software development and healthcare. However, this deepening trust introduces a novel attack surface: Agent-Mediated Deception (AMD), where compromised agents are weaponized against their human users. While extensive research focuses on agent-centric threats, human susceptibility to deception by a compromised agent remains unexplored. We present the first large-scale empirical study with 303 participants to measure human susceptibility to AMD. This is based on HAT-Lab (Human-Agent Trust Laboratory), a high-fidelity research platform we develop, featuring nine carefully crafted scenarios spanning everyday and professional domains (e.g., healthcare, software development, human resources). Our 10 key findings reveal significant vulnerabilities and provide future defense perspectives. Specifically, only 8.6% of participants perceive AMD attacks, while domain experts show increased susceptibility in certain scenarios. We identify six cognitive failure modes in users and find that their risk awareness often fails to translate to protective behavior. The defense analysis reveals that effective warnings should interrupt workflows with low verification costs. With experiential learning based on HAT-Lab, over 90% of users who perceive risks report increased caution against AMD. This work provides empirical evidence and a platform for human-centric agent security research.

The integration of external data services (e.g., Model Context Protocol, MCP) has made large language model-based agents increasingly powerful for complex task execution. However, this advancement introduces critical security vulnerabilities, particularly indirect prompt injection (IPI) attacks. Existing attack methods are limited by their reliance on static patterns and evaluation on simple language models, failing to address the fast-evolving nature of modern AI agents. We introduce AdapTools, a novel adaptive IPI attack framework that selects stealthier attack tools and generates adaptive attack prompts to create a rigorous security evaluation environment. Our approach comprises two key components: (1) Adaptive Attack Strategy Construction, which develops transferable adversarial strategies for prompt optimization, and (2) Attack Enhancement, which identifies stealthy tools capable of circumventing task-relevance defenses. Comprehensive experimental evaluation shows that AdapTools achieves a 2.13 times improvement in attack success rate while degrading system utility by a factor of 1.78. Notably, the framework maintains its effectiveness even against state-of-the-art defense mechanisms. Our method advances the understanding of IPI attacks and provides a useful reference for future research.

Large Language Model (LLM) agents are susceptible to Indirect Prompt Injection (IPI) attacks, where malicious instructions in retrieved content hijack the agent's execution. Existing defenses typically rely on strict filtering or refusal mechanisms, which suffer from a critical limitation: over-refusal, prematurely terminating valid agentic workflows. We propose ICON, a probing-to-mitigation framework that neutralizes attacks while preserving task continuity. Our key insight is that IPI attacks leave distinct over-focusing signatures in the latent space. We introduce a Latent Space Trace Prober to detect attacks based on high intensity scores. Subsequently, a Mitigating Rectifier performs surgical attention steering that selectively manipulate adversarial query key dependencies while amplifying task relevant elements to restore the LLM's functional trajectory. Extensive evaluations on multiple backbones show that ICON achieves a competitive 0.4% ASR, matching commercial grade detectors, while yielding a over 50% task utility gain. Furthermore, ICON demonstrates robust Out of Distribution(OOD) generalization and extends effectively to multi-modal agents, establishing a superior balance between security and efficiency.

We report an exploratory red-teaming study of autonomous language-model-powered agents deployed in a live laboratory environment with persistent memory, email accounts, Discord access, file systems, and shell execution. Over a two-week period, twenty AI researchers interacted with the agents under benign and adversarial conditions. Focusing on failures emerging from the integration of language models with autonomy, tool use, and multi-party communication, we document eleven representative case studies. Observed behaviors include unauthorized compliance with non-owners, disclosure of sensitive information, execution of destructive system-level actions, denial-of-service conditions, uncontrolled resource consumption, identity spoofing vulnerabilities, cross-agent propagation of unsafe practices, and partial system takeover. In several cases, agents reported task completion while the underlying system state contradicted those reports. We also report on some of the failed attempts. Our findings establish the existence of security-, privacy-, and governance-relevant vulnerabilities in realistic deployment settings. These behaviors raise unresolved questions regarding accountability, delegated authority, and responsibility for downstream harms, and warrant urgent attention from legal scholars, policymakers, and researchers across disciplines. This report serves as an initial empirical contribution to that broader conversation.

Agentic systems based on large language models (LLMs) operate not merely as text generators but as autonomous entities that dynamically retrieve information and invoke tools. This execution model shifts the attack surface from traditional build-time artifacts to inference-time dependencies, exposing agents to manipulation through untrusted data and probabilistic capability resolution. While prior work has examined model-level vulnerabilities, security risks arising from the complex, cyclic runtime behavior of agents remain fragmented. This paper systematizes existing research into a unified runtime framework. We categorize threats into data supply chain attacks (distinguishing between transient context injection and persistent memory poisoning) and tool supply chain attacks (spanning discovery, implementation, and invocation phases). Crucially, we identify the emergence of the Viral Agent Loop, where agents effectively become vectors for self-propagating generative worms that require no code vulnerabilities to spread. We argue for a transition to a Zero-Trust Runtime Architecture, where context is treated as untrusted control flow, and tool execution is bounded by cryptographic provenance rather than semantic likelihood.

Agent hijacking, highlighted by OWASP as a critical threat to the Large Language Model (LLM) ecosystem, enables adversaries to manipulate execution by injecting malicious instructions into retrieved content. Most existing attacks rely on manually crafted, semantics-driven prompt manipulation, which often yields low attack success rates and limited transferability to closed-source commercial models. In this paper, we propose Phantom, an automated agent hijacking framework built upon Structured Template Injection that targets the fundamental architectural mechanisms of LLM agents. Our key insight is that agents rely on specific chat template tokens to separate system, user, assistant, and tool instructions. By injecting optimized structured templates into the retrieved context, we induce role confusion and cause the agent to misinterpret the injected content as legitimate user instructions or prior tool outputs. To enhance attack transferability against black-box agents, Phantom introduces a novel attack template search framework. We first perform multi-level template augmentation to increase structural diversity and then train a Template Autoencoder (TAE) to embed discrete templates into a continuous, searchable latent space. Subsequently, we apply Bayesian optimization to efficiently identify optimal adversarial vectors that are decoded into high-potency structured templates. Extensive experiments on Qwen, GPT, and Gemini demonstrate that our framework significantly outperforms existing baselines in both Attack Success Rate (ASR) and query efficiency. Moreover, we identified over 70 vulnerabilities in real-world commercial products that have been confirmed by vendors, underscoring the practical severity of structured template-based hijacking and providing an empirical foundation for securing next-generation agentic systems.

Model Context Protocols (MCPs) provide a unified platform for agent systems to discover, select, and orchestrate tools across heterogeneous execution environments. As MCP-based systems scale to incorporate larger tool catalogs and multiple concurrently connected MCP servers, traditional tool-by-tool invocation increases coordination overhead, fragments state management, and limits support for wide-context operations. To address these scalability challenges, recent MCP designs have incorporated code execution as a first-class capability, an approach called Code Execution MCP (CE-MCP). This enables agents to consolidate complex workflows, such as SQL querying, file analysis, and multi-step data transformations, into a single program that executes within an isolated runtime environment. In this work, we formalize the architectural distinction between context-coupled (traditional) and context-decoupled (CE-MCP) models, analyzing their fundamental scalability trade-offs. Using the MCP-Bench framework across 10 representative servers, we empirically evaluate task behavior, tool utilization patterns, execution latency, and protocol efficiency as the scale of connected MCP servers and available tools increases, demonstrating that while CE-MCP significantly reduces token usage and execution latency, it introduces a vastly expanded attack surface. We address this security gap by applying the MAESTRO framework, identifying sixteen attack classes across five execution phases-including specific code execution threats such as exception-mediated code injection and unsafe capability synthesis. We validate these vulnerabilities through adversarial scenarios across multiple LLMs and propose a layered defense architecture comprising containerized sandboxing and semantic gating. Our findings provide a rigorous roadmap for balancing scalability and security in production-ready executable agent workflows.

Clawdbot is a self-hosted, tool-using personal AI agent with a broad action space spanning local execution and web-mediated workflows, which raises heightened safety and security concerns under ambiguity and adversarial steering. We present a trajectory-centric evaluation of Clawdbot across six risk dimensions. Our test suite samples and lightly adapts scenarios from prior agent-safety benchmarks (including ATBench and LPS-Bench) and supplements them with hand-designed cases tailored to Clawdbot's tool surface. We log complete interaction trajectories (messages, actions, tool-call arguments/outputs) and assess safety using both an automated trajectory judge (AgentDoG-Qwen3-4B) and human review. Across 34 canonical cases, we find a non-uniform safety profile: performance is generally consistent on reliability-focused tasks, while most failures arise under underspecified intent, open-ended goals, or benign-seeming jailbreak prompts, where minor misinterpretations can escalate into higher-impact tool actions. We supplemented the overall results with representative case studies and summarized the commonalities of these cases, analyzing the security vulnerabilities and typical failure modes that Clawdbot is prone to trigger in practice.

The rapid evolution of GUI-enabled agents has rendered traditional CAPTCHAs obsolete. While previous benchmarks like OpenCaptchaWorld established a baseline for evaluating multimodal agents, recent advancements in reasoning-heavy models, such as Gemini3-Pro-High and GPT-5.2-Xhigh have effectively collapsed this security barrier, achieving pass rates as high as 90% on complex logic puzzles like "Bingo". In response, we introduce Next-Gen CAPTCHAs, a scalable defense framework designed to secure the next-generation web against the advanced agents. Unlike static datasets, our benchmark is built upon a robust data generation pipeline, allowing for large-scale and easily scalable evaluations, notably, for backend-supported types, our system is capable of generating effectively unbounded CAPTCHA instances. We exploit the persistent human-agent "Cognitive Gap" in interactive perception, memory, decision-making, and action. By engineering dynamic tasks that require adaptive intuition rather than granular planning, we re-establish a robust distinction between biological users and artificial agents, offering a scalable and diverse defense mechanism for the agentic era.