Autonomous agent frameworks built upon large language models (LLMs) are evolving into complex, tool-integrated, and continuously operating systems, introducing security risks beyond traditional prompt-level vulnerabilities. As this paradigm is still at an early stage of development, a timely and systematic understanding of its security implications is increasingly important. Although a growing body of work has examined different attack surfaces and defense problems in agent systems, existing studies remain scattered across individual aspects of agent security, and there is still a lack of a layered review on this topic. To address this gap, this survey presents a layered review of security risks and defense strategies in autonomous agent frameworks, with OpenClaw as a case study. We organize the analysis into four security-relevant layers: the context and instruction layer, the tool and action layer, the state and persistence layer, and the ecosystem and automation layer. For each layer, we summarize its functional role, representative security risks, and corresponding defense strategies. Based on this layered analysis, we further identify that threats in autonomous agent frameworks may propagate across layers, from manipulated inputs to unsafe actions, persistent state contamination, and broader ecosystem-level impact. Finally, we highlight potential key challenges, including research imbalance across layers, the lack of long-horizon evaluation, and weak ecosystem trust models, and outline future directions toward more systematic and integrated defenses.

Alignment faking (AF) occurs when an LLM strategically complies with training objectives to avoid value modification, reverting to prior preferences once monitoring is lifted. Current detection methods focus on conversational settings and rely primarily on Chain-of-Thought (CoT) analysis, which provides a reliable signal when strategic reasoning surfaces, but cannot distinguish deception from capability failures if traces are absent or unfaithful. We formalize AF as a composite behavioural event and detect it through observable tool selection, where the LLM selects the safe tool when unmonitored, but switches to the unsafe tool under monitoring that rewards helpfulness over safety, while its reasoning still acknowledges the safe choice. We release a dataset of 108 enterprise IT scenarios spanning Security, Privacy, and Integrity domains under Corruption and Sabotage pressures. Evaluating six frontier LLMs across five independent runs, we find mean AF detection rates between 3.5% and 23.7%, with vulnerability profiles varying by domain and pressure type. These results suggest that susceptibility reflects training methodology rather than capability alone.

Structured-workflow agents driven by large language models execute tool calls against sensitive external environments. We propose \codename, a telemetry-driven behavioral anomaly detection firewall. Drawing on sequence-based intrusion detection, \codename\ compiles verified benign tool-call telemetry into a parameterized deterministic finite automaton (pDFA). The model defines permitted tool sequences, sequential contexts, and parameter bounds. At runtime, a lightweight gateway enforces these boundaries via an $O(1)$ state-transition structural lookup, shifting computationally expensive analysis entirely offline. Evaluated on the Agent Security Bench (ASB), \codename\ achieves a 5.6\% macro-averaged attack success rate (ASR) across five scenarios. Within three structured workflows, ASR drops to 2.2\%, outperforming Aegis, a state-of-the-art stateless scanner, at 12.8\%. \codename\ achieves 0\% ASR on multi-step and context-sequential attacks in structured settings. Furthermore, against 1,000 algorithmically spliced exfiltration payloads, only 1.4\% matched valid structural paths, all of which failed end-to-end string parameter guards (0 successes out of 14 surviving paths, 95\% CI [0\%, 23.2\%]). \codename\ introduces just 2.2~ms of per-call latency (a 3.7$\times$ speedup over \textsc{Aegis}) while maintaining a 2.0\% benign task failure rate (BTFR) on benign workloads. Modeling the behavioral trajectory effectively collapses the available attack surface, but unmaintained continuous parameter bounds remain vulnerable to synonym-substitution attacks (18\% evasion rate). Thus, exact-match whitelisting of sensitive parameters ultimately bears the final defensive load against execution.

Web agents have emerged as an effective paradigm for automating interactions with complex web environments, yet remain vulnerable to prompt injection attacks that embed malicious instructions into webpage content to induce unintended actions. This threat is further amplified for screenshot-based web agents, which operate on rendered visual webpages rather than structured textual representations, making predominant text-centric defenses ineffective. Although multimodal detection methods have been explored, they often rely on large vision-language models (VLMs), incurring significant computational overhead. The bottleneck lies in the complexity of modern webpages: VLMs must comprehend the global semantics of an entire page, resulting in substantial inference time and GPU memory usage. This raises a critical question: can we detect prompt injection attacks from screenshots in a lightweight manner? In this paper, we observe that injected webpages exhibit distinct characteristics compared to benign ones from both visual and textual perspectives. Building on this insight, we propose SnapGuard, a lightweight yet accurate method that reformulates prompt injection detection as multimodal representation analysis over webpage screenshots. SnapGuard leverages two complementary signals: a visual stability indicator that identifies abnormally smooth gradient distributions induced by malicious content, and action-oriented textual signals recovered via contrast-polarity reversal. Extensive evaluations across eight attacks and two benign settings demonstrate that SnapGuard achieves an F1 score of 0.75, outperforming GPT-4o-prompt while being 8x faster (1.81s vs. 14.50s) and introducing no additional memory overhead.

Autonomous AI agents deployed on platforms such as OpenClaw face prompt injection, memory poisoning, supply-chain attacks, and social engineering, yet existing defences address only the platform perimeter, leaving the agent's own threat judgement entirely untrained. We present ClawdGo, a framework for endogenous security awareness training: we teach the agent to recognise and reason about threats from the inside, at inference time, with no model modification. Four contributions are introduced: TLDT (Three-Layer Domain Taxonomy) organises 12 trainable dimensions across Self-Defence, Owner-Protection, and Enterprise-Security layers; ASAT (Autonomous Security Awareness Training) is a self-play loop where the agent alternates attacker, defender, and evaluator roles under weakest-first curriculum scheduling; CSMA (Cross-Session Memory Accumulation) compounds skill gains via a four-layer persistent memory architecture and Axiom Crystallisation Promotion (ACP); and SACP (Security Awareness Calibration Problem) formalises the precision-recall tradeoff introduced by endogenous training. Live experiments show weakest-first ASAT raises average TLDT score from 80.9 to 96.9 over 16 sessions, outperforming uniform-random scheduling by 6.5 points and covering 11 of 12 dimensions. CSMA retains the full gain across sessions; cold-start ablation recovers only 2.4 points, leaving a 13.6-point gap. E-mode generates 32 TLDT-conformant scenarios covering all 12 dimensions. SACP is observed when a heavily trained agent classifies a legitimate capability assessment as prompt injection (30/160).

Large Language Model (LLM) agents are increasingly used to automate complex workflows, but integrating untrusted external data with privileged execution exposes them to severe security risks, particularly direct and indirect prompt injection. Existing defenses face significant challenges in balancing security with utility, often encountering a trade-off where rigorous protection leads to over-defense, or where subtle indirect injections bypass detection. Drawing inspiration from operating system virtualization, we propose AgentVisor, a novel defense framework that enforces semantic privilege separation. AgentVisor treats the target agent as an untrusted guest and intercepts tool calls via a trusted semantic visor. Central to our approach is a rigorous audit protocol grounded in classic OS security primitives, designed to systematically mitigate both direct and indirect injection attacks. Furthermore, we introduce a one-shot self-correction mechanism that transforms security violations into constructive feedback, enabling agents to recover from attacks. Extensive experiments show that AgentVisor reduces the attack success rate to 0.65%, achieving this strong defense while incurring only a 1.45% average decrease in utility relative to the No Defense scenario, demonstrating superior performance compared to existing defense methods.

The rapid integration of Large Language Models (LLMs) into Multi-Agent Systems (MAS) has significantly enhanced their collaborative problem-solving capabilities, but it has also expanded their attack surfaces, exposing them to vulnerabilities such as prompt infection and compromised inter-agent communication. While emerging graph-based anomaly detection methods show promise in protecting these networks, the field currently lacks a standardized, reproducible environment to train these models and evaluate their efficacy. To address this gap, we introduce Gammaf (Graph-based Anomaly Monitoring for LLM Multi-Agent systems Framework), an open-source benchmarking platform. Gammaf is not a novel defense mechanism itself, but rather a comprehensive evaluation architecture designed to generate synthetic multi-agent interaction datasets and benchmark the performance of existing and future defense models. The proposed framework operates through two interdependent pipelines: a Training Data Generation stage, which simulates debates across varied network topologies to capture interactions as robust attributed graphs, and a Defense System Benchmarking stage, which actively evaluates defense models by dynamically isolating flagged adversarial nodes during live inference rounds. Through rigorous evaluation using established defense baselines (XG-Guard and BlindGuard) across multiple knowledge tasks (such as MMLU-Pro and GSM8K), we demonstrate Gammaf's high utility, topological scalability, and execution efficiency. Furthermore, our experimental results reveal that equipping an LLM-MAS with effective attack remediation not only recovers system integrity but also substantially reduces overall operational costs by facilitating early consensus and cutting off the extensive token generation typical of adversarial agents.

Agentic AI systems face security challenges that stateless large language models do not. They plan across extended horizons, maintain persistent memory, invoke external tools, and coordinate with peer agents. Existing security analyses organize threats by attack type (prompt injection, jailbreaking), but provide no principled model of which architectural component is vulnerable or over what timescale the threat manifests. This paper makes five contributions. First, we introduce the Layered Attack Surface Model (LASM), a seven-layer framework that maps threats to distinct architectural components: Foundation, Cognitive, Memory, Tool Execution, Multi-Agent Coordination, Ecosystem, and Governance, the accountability and observability layer that spans the stack analogously to the network management plane. Second, we introduce attack temporality as an orthogonal analytical dimension with four classes: Instantaneous (T1), Session-Persistent (T2), Cross-Session Cumulative (T3), and Sub-Session-Stack, Non-Session-Bounded (T4). Third, through a systematic review of 94 papers (2021--2025), we show that the most dangerous emerging threats concentrate at the intersection of high-layer attacks (L5--L7) and slow-burn temporality (T3--T4): covert agent collusion, long-term memory poisoning, MCP supply-chain compromise, and alignment failure that manifests as an insider threat with no external adversary. Only 8 of 120 paper-cell assignments (7%) fall in this zone. Fourth, we propose a cross-layer defense taxonomy spanning all seven LASM layers and all four temporality classes, exposing which threat classes existing defenses leave unaddressed. Fifth, we survey evaluation benchmarks, identify five research gaps in the under-studied high-layer, slow-burn zone, and argue that agentic security must be treated as a distributed systems problem embedded in an adversarial ecosystem.

Emerging AR-LLM-based Social Engineering attack (e.g., SEAR) is at the edge of posing great threats to real-world social life. In such AR-LLM-SE attack, the attacker can leverage AR (Augmented Reality) glass to capture the image and vocal information of the target, using the LLM to identify the target and generate the social profile, using the LLM agents to apply social engineering strategies for conversation suggestion to win the target trust and perform phishing afterwards. Current defensive approaches, such as role-based access control or data flow tracking, are not directly applicable to the convergent AR-LLM ecosystem (considering embedded AR device and opaque LLM inference), leaving an emerging and potent social engineering threat that existing privacy paradigms are ill-equipped to address. This necessitates a shift beyond solely human-centric measures like legislation and user education toward enforceable vendor policies and platform-level restrictions. Realizing this vision, however, faces significant technical challenges: securing resource-constrained AR-embedded devices, implementing fine-grained access control within opaque LLM inferences, and governing adaptive interactive agents. To address these challenges, we present UNSEEN, a coordinated cross-stack defense that combines an AR ACL (Access Control Layer) for identity-gated sensing, F-RMU-based LLM unlearning for sensitive profile suppression, and runtime agent guardrails for adaptive interaction control. We evaluate UNSEEN in an IRB-approved user study with 60 participants and a dataset of 360 annotated conversations across realistic social scenarios.

Agent skills introduce a new and more severe form of indirect injection for LLM agents: unlike traditional indirect prompt injection, attackers can hide malicious instructions inside a dense, action-oriented skill that already functions as a legitimate instruction source. We study pre-execution skill-poison detection and show that successful skill poisoning induces a structured internal effect, attention hijacking, in which response-time attention shifts from trusted context to malicious skill spans and drives harmful behavior. Motivated by this mechanism, we propose RouteGuard, a frozen-backbone detector that combines response-conditioned attention and hidden-state alignment through reliability-gated late fusion. Across both real and synthetic open-source skill benchmarks, RouteGuard is consistently the strongest or most robust detector; on the critical Skill-Inject channel slice, it reaches 0.8834 F1 and recovers 90.51% of description attacks missed by lexical screening, showing that defending against skill poisoning requires internal-signal detection rather than text-only filtering

The offensive security landscape is highly fragmented: enterprise platforms avoid memory-corruption vulnerabilities due to Denial of Service (DoS) risks, Automatic Exploit Generation (AEG) systems suffer from semantic blindness, and Large Language Model (LLM) agents face safety alignment filters and "Live Fire" execution hazards. We introduce Automation-Exploit, a fully autonomous Multi-Agent System (MAS) framework designed for adaptive offensive security in complex black-box scenarios. It bridges the abstraction gap between reconnaissance and exploitation by autonomously exfiltrating executables and contextual intelligence across multiple protocols, using this data to fuel both logical and binary attack chains. The framework introduces an adaptive safety architecture to mitigate DoS risks. While it natively resolves logical and web-based vulnerabilities, it employs a conditional isomorphic validation for high-risk memory-corruption flaws: if the target binary is successfully exfiltrated, it dynamically instantiates a cross-platform digital twin. By enforcing strict state synchronization, including libc alignment and runtime file descriptor hooking, potentially destructive payloads are iteratively debugged in an isolated replica. This enables a highly risk-mitigated "one-shot" execution on the physical target. Empirical evaluations across eight scenarios, including undocumented zero-day environments to rule out LLM data contamination, validate the framework's architectural resilience, demonstrating its ability to prevent "live fire" crashes and execute risk-mitigated compromises on actual targets.

The growth of agentic AI has drawn significant attention to function calling Large Language Models (LLMs), which are designed to extend the capabilities of AI-powered system by invoking external functions. Injection and jailbreaking attacks have been extensively explored to showcase the vulnerabilities of LLMs to user prompt manipulation. The expanded capabilities of agentic models introduce further vulnerabilities via their function calling interface. Recent work in LLM security showed that function calling can be abused, leading to data tampering and theft, causing disruptive behavior such as endless loops, or causing LLMs to produce harmful content in the style of jailbreaking attacks. This paper introduces a novel function hijacking attack (FHA) that manipulates the tool selection process of agentic models to force the invocation of a specific, attacker-chosen function. While existing attacks focus on semantic preference of the model for function-calling tasks, we show that FHA is largely agnostic to the context semantics and robust to the function sets, making it applicable across diverse domains. We further demonstrate that FHA can be trained to produce universal adversarial functions, enabling a single attacked function to hijack tool selection across multiple queries and payload configurations. We conducted experiments on 5 different models, including instructed and reasoning variants, reaching 70% to 100% ASR over the established BFCL dataset. Our findings further demonstrate the need for strong guardrails and security modules for agentic systems.

AI-agent guardrails are memoryless: each message is judged in isolation, so an adversary who spreads a single attack across dozens of sessions slips past every session-bound detector because only the aggregate carries the payload. We make three contributions to cross-session threat detection. (1) Dataset. CSTM-Bench is 26 executable attack taxonomies classified by kill-chain stage and cross-session operation (accumulate, compose, launder, inject_on_reader), each bound to one of seven identity anchors that ground-truth "violation" as a policy predicate, plus matched Benign-pristine and Benign-hard confounders. Released on Hugging Face as intrinsec-ai/cstm-bench with two 54-scenario splits: dilution (compositional) and cross_session (12 isolation-invisible scenarios produced by a closed-loop rewriter that softens surface phrasing while preserving cross-session artefacts). (2) Measurement. Framing cross-session detection as an information bottleneck to a downstream correlator LLM, we find that a session-bound judge and a Full-Log Correlator concatenating every prompt into one long-context call both lose roughly half their attack recall moving from dilution to cross_session, well inside any frontier context window. Scope: 54 scenarios per shard, one correlator family (Anthropic Claude), no prompt optimisation; we release it to motivate larger, multi-provider datasets. (3) Algorithm and metric. A bounded-memory Coreset Memory Reader retaining highest-signal fragments at $K=50$ is the only reader whose recall survives both shards. Because ranker reshuffles break KV-cache prefix reuse, we promote $\mathrm{CSR\_prefix}$ (ordered prefix stability, LLM-free) to a first-class metric and fuse it with detection into $\mathrm{CSTM} = 0.7 F_1(\mathrm{CSDA@action}, \mathrm{precision}) + 0.3 \mathrm{CSR\_prefix}$, benchmarking rankers on a single Pareto of recall versus serving stability.

AI agents promise to serve as general-purpose personal assistants for their users, which requires them to have access to private user data (e.g., personal and financial information). This poses a serious risk to security and privacy. Adversaries may attack the AI model (e.g., via prompt injection) to exfiltrate user data. Furthermore, sharing private data with an AI agent requires users to trust a potentially unscrupulous or compromised AI model provider with their private data. This paper presents GAAP (Guaranteed Accounting for Agent Privacy), an execution environment for AI agents that guarantees confidentiality for private user data. Through dynamic and directed user prompts, GAAP collects permission specifications from users describing how their private data may be shared, and GAAP enforces that the agent's disclosures of private user data, including disclosures to the AI model and its provider, comply with these specifications. Crucially, GAAP provides this guarantee deterministically, without trusting the agent with private user data, and without requiring any AI model or the user prompt to be free of attacks. GAAP enforces the user's permission specification by tracking how the AI agent accesses and uses private user data. It augments Information Flow Control with novel persistent data stores and annotations that enable it to track the flow of private information both across execution steps within a single task, and also over multiple tasks separated in time. Our evaluation confirms that GAAP blocks all data disclosure attacks, including those that make other state-of-the-art systems disclose private user data to untrusted parties, without a significant impact on agent utility.

Recent advances in embodied Vision-Language Agentic Systems (VLAS), powered by large vision-language models (LVLMs), enable AI systems to perceive and reason over real-world scenes. Within this context, environmental signals such as traffic lights are essential in-band signals that can and should influence agent behavior. However, similar signals could also be crafted to operate as misleading visual injections, overriding user intent and posing security risks. This duality creates a fundamental challenge: agents must respond to legitimate environmental cues while remaining robust to misleading ones. We refer to this tension as trust boundary confusion. To study this behavior, we design a dual-intent dataset and evaluation framework, through which we show that current LVLM-based agents fail to reliably balance this trade-off, either ignoring useful signals or following harmful ones. We systematically evaluate 7 LVLM agents across multiple embodied settings under both structure-based and noise-based visual injections. To address these vulnerabilities, we propose a multi-agent defense framework that separates perception from decision-making to dynamically assess the reliability of visual inputs. Our approach significantly reduces misleading behaviors while preserving correct responses and provides robustness guarantees under adversarial perturbations. The code of the evaluation framework and artifacts are made available at https://anonymous.4open.science/r/Visual-Prompt-Inject.

The transition of agentic AI from brittle prototypes to production systems is stalled by a pervasive crisis of craft. We suggest that the prevailing orchestration paradigm-delegating the system control loop to large language models and merely patching with heuristic guardrails-is the root cause of this fragility. Instead, we propose Arbiter-K, a Governance-First execution architecture that reconceptualizes the underlying model as a Probabilistic Processing Unit encapsulated by a deterministic, neuro-symbolic kernel. Arbiter-K implements a Semantic Instruction Set Architecture (ISA) to reify probabilistic messages into discrete instructions. This allows the kernel to maintain a Security Context Registry and construct an Instruction Dependency Graph at runtime, enabling active taint propagation based on the data-flow pedigree of each reasoning node. By leveraging this mechanism, Arbiter-K precisely interdicts unsafe trajectories at deterministic sinks (e.g., high-risk tool calls or unauthorized network egress) and enables autonomous execution correction and architectural rollback when security policies are triggered. Evaluations on OpenClaw and NanoBot demonstrate that Arbiter-K enforces security as a microarchitectural property, achieving 76% to 95% unsafe interception for a 92.79% absolute gain over native policies. The code is publicly available at https://github.com/cure-lab/ArbiterOS.

Large Language Model (LLM) agents provide powerful automation capabilities, but they also create a substantially broader attack surface than traditional applications due to their tight integration with non-deterministic models and third-party services. While current deployments primarily rely on cloud-hosted services, emerging designs increasingly execute agents directly on edge devices to reduce latency and enhance user privacy. However, securely hosting such complex agent pipelines on edge devices remains challenging. These deployments must protect proprietary assets (e.g., system prompts and model weights) and sensitive runtime state on heterogeneous platforms that are vulnerable to software attacks and potentially controlled by malicious users. To address these challenges, we present AgenTEE, a system for deploying confidential agent pipelines on edge devices. AgenTEE places the agent runtime, inference engine, and third-party applications into independently attested confidential virtual machines (cVMs) and mediates their interaction through explicit, verifiable communication channels. Built on Arm Confidential Compute Architecture (CCA), a recent extension to Arm platforms, AgenTEE enforces strong system-level isolation of sensitive assets and runtime state. Our evaluation shows that such multi-cVMs system is practical, achieving near-native performance with less than 5.15% runtime overhead compared to commodity OS multi-process deployments.

Existing AI agent safety benchmarks focus on generic criminal harm (cybercrime, harassment, weapon synthesis), leaving a systematic blind spot for a distinct and commercially consequential threat category: agents harming their own deployers. Real-world incidents illustrate the gap: Slack AI credential exfiltration (Aug 2024), Microsoft 365 Copilot calendar-injection leaks (Jan 2024), and a Meta agent unauthorized forum post exposing operational data (Mar 2026). We propose Owner-Harm, a formal threat model with eight categories of agent behavior damaging the deployer. We quantify the defense gap on two benchmarks: a compositional safety system achieves 100% TPR / 0% FPR on AgentHarm (generic criminal harm) yet only 14.8% (4/27; 95% CI: 5.9%-32.5%) on AgentDojo injection tasks (prompt-injection-mediated owner harm). A controlled generic-LLM baseline shows the gap is not inherent to owner-harm (62.7% vs. 59.3%, delta 3.4 pp) but arises from environment-bound symbolic rules that fail to generalize across tool vocabularies. On a post-hoc 300-scenario owner-harm benchmark, the gate alone achieves 75.3% TPR / 3.3% FPR; adding a deterministic post-audit verifier raises overall TPR to 85.3% (+10.0 pp) and Hijacking detection from 43.3% to 93.3%, demonstrating strong layer complementarity. We introduce the Symbolic-Semantic Defense Generalization (SSDG) framework relating information coverage to detection rate. Two SSDG experiments partially validate it: context deprivation amplifies the detection gap 3.4x (R = 3.60 vs. R = 1.06); context injection reveals structured goal-action alignment, not text concatenation, is required for effective owner-harm detection.

Large language model (LLM) agents are vulnerable to prompt-injection attacks that propagate through multi-step workflows, tool interactions, and persistent context, making input-output filtering alone insufficient for reliable protection. This paper presents SafeAgent, a runtime security architecture that treats agent safety as a stateful decision problem over evolving interaction trajectories. The proposed design separates execution governance from semantic risk reasoning through two coordinated components: a runtime controller that mediates actions around the agent loop and a context-aware decision core that operates over persistent session state. The core is formalized as a context-aware advanced machine intelligence and instantiated through operators for risk encoding, utility-cost evaluation, consequence modeling, policy arbitration, and state synchronization. Experiments on Agent Security Bench (ASB) and InjecAgent show that SafeAgent consistently improves robustness over baseline and text-level guardrail methods while maintaining competitive benign-task performance. Ablation studies further show that recovery confidence and policy weighting determine distinct safety-utility operating points.

The evolution from static ranking models to Agentic Recommender Systems (Agentic RecSys) empowers AI agents to maintain long-term user profiles and autonomously plan service tasks. While this paradigm shift enhances personalization, it introduces a vulnerability: reliance on Long-term Memory (LTM). In this paper, we uncover a threat termed "Visual Inception." Unlike traditional adversarial attacks that seek immediate misclassification, Visual Inception injects triggers into user-uploaded images (e.g., lifestyle photos) that act as "sleeper agents" within the system's memory. When retrieved during future planning, these poisoned memories hijack the agent's reasoning chain, steering it toward adversary-defined goals (e.g., promoting high-margin products) without prompt injection. To mitigate this, we propose CognitiveGuard, a dual-process defense framework inspired by human cognition. It consists of a System 1 Perceptual Sanitizer (diffusion-based purification) to cleanse sensory inputs and a System 2 Reasoning Verifier (counterfactual consistency checks) to detect anomalies in memory-driven planning. Extensive experiments on a mock e-commerce agent environment demonstrate that Visual Inception achieves about 85% Goal-Hit Rate (GHR), while CognitiveGuard reduces this risk to around 10% with configurable latency trade-offs (about 1.5s in lite mode to about 6.5s for full sequential verification), without quality degradation under our setup.