Large language models (LLMs) have evolved from simple chatbots into autonomous agents capable of performing complex tasks such as editing production code, orchestrating workflows, and taking higher-stakes actions based on untrusted inputs like webpages and emails. These capabilities introduce new security risks that existing security measures, such as model fine-tuning or chatbot-focused guardrails, do not fully address. Given the higher stakes and the absence of deterministic solutions to mitigate these risks, there is a critical need for a real-time guardrail monitor to serve as a final layer of defense, and support system level, use case specific safety policy definition and enforcement. We introduce LlamaFirewall, an open-source security focused guardrail framework designed to serve as a final layer of defense against security risks associated with AI Agents. Our framework mitigates risks such as prompt injection, agent misalignment, and insecure code risks through three powerful guardrails: PromptGuard 2, a universal jailbreak detector that demonstrates clear state of the art performance; Agent Alignment Checks, a chain-of-thought auditor that inspects agent reasoning for prompt injection and goal misalignment, which, while still experimental, shows stronger efficacy at preventing indirect injections in general scenarios than previously proposed approaches; and CodeShield, an online static analysis engine that is both fast and extensible, aimed at preventing the generation of insecure or dangerous code by coding agents. Additionally, we include easy-to-use customizable scanners that make it possible for any developer who can write a regular expression or an LLM prompt to quickly update an agent's security guardrails.

In this paper, we introduce OSUniverse: a benchmark of complex, multimodal desktop-oriented tasks for advanced GUI-navigation AI agents that focuses on ease of use, extensibility, comprehensive coverage of test cases, and automated validation. We divide the tasks in increasing levels of complexity, from basic precision clicking to multistep, multiapplication tests requiring dexterity, precision, and clear thinking from the agent. In version one of the benchmark, presented here, we have calibrated the complexity of the benchmark test cases to ensure that the SOTA (State of the Art) agents (at the time of publication) do not achieve results higher than 50%, while the average white collar worker can perform all these tasks with perfect accuracy. The benchmark can be scored manually, but we also introduce an automated validation mechanism that has an average error rate less than 2%. Therefore, this benchmark presents solid ground for fully automated measuring of progress, capabilities and the effectiveness of GUI-navigation AI agents over the short and medium-term horizon. The source code of the benchmark is available at https://github.com/agentsea/osuniverse.

--This study explores the application of chaos engineering to enhance the robustness of Large Language Model-Based Multi-Agent Systems (LLM-MAS) in production-like environments under real-world conditions. LLM-MAS can potentially improve a wide range of tasks, from answering questions and generating content to automating customer support and improving decision-making processes. However, LLM-MAS in production or preproduction environments can be vulnerable to emergent errors or disruptions, such as hallucinations, agent failures, and agent communication failures. This study proposes a chaos engineering framework to proactively identify such vulnerabilities in LLM-MAS, assess and build resilience against them, and ensure reliable performance in critical applications. I NTRODUCTION Large Language Models (LLMs) such as Bing [1], Gemini [2], and ChatGPT [3] have transformed natural language processing (NLP) through innovations such as transformer architectures [4] and large-scale pretraining [5].

In this paper, we present a global-to-local task-aware fault detection and identification algorithm to detect failures in a multi-spacecraft system performing a collaborative inspection (referred to as global) task. The inspection task is encoded as a cost functional $\costH$ that informs global (task allocation and assignment) and local (agent-level) decision-making. The metric $\costH$ is a function of the inspection sensor model, and the agent full-pose. We use the cost functional $\costH$ to design a metric that compares the expected and actual performance to detect the faulty agent using a threshold. We use higher-order cost gradients $\costH$ to derive a new metric to identify the type of fault, including task-specific sensor fault, an agent-level actuator, and sensor faults. Furthermore, we propose an approach to design adaptive thresholds for each fault mentioned above to incorporate the time dependence of the inspection task. We demonstrate the efficacy of the proposed method empirically, by simulating and detecting faults (such as inspection sensor faults, actuators, and sensor faults) in a low-Earth orbit collaborative spacecraft inspection task using the metrics and the threshold designed using the global task cost $\costH$.

As language models (LMs) are increasingly deployed as autonomous agents, their robust adherence to human-assigned objectives becomes crucial for safe operation. When these agents operate independently for extended periods without human oversight, even initially well-specified goals may gradually shift. Detecting and measuring goal drift - an agent's tendency to deviate from its original objective over time - presents significant challenges, as goals can shift gradually, causing only subtle behavioral changes. This paper proposes a novel approach to analyzing goal drift in LM agents. In our experiments, agents are first explicitly given a goal through their system prompt, then exposed to competing objectives through environmental pressures. We demonstrate that while the best-performing agent (a scaffolded version of Claude 3.5 Sonnet) maintains nearly perfect goal adherence for more than 100,000 tokens in our most difficult evaluation setting, all evaluated models exhibit some degree of goal drift. We also find that goal drift correlates with models' increasing susceptibility to pattern-matching behaviors as the context length grows.

This paper explores the emergence of language in multi-agent reinforcement learning (MARL) using transformers. Existing methods such as RIAL, DIAL, and CommNet enable agent communication but lack interpretability. We propose Differentiable Inter-Agent Transformers (DIAT), which leverage self-attention to learn symbolic, human-understandable communication protocols. Through experiments, DIAT demonstrates the ability to encode observations into interpretable vocabularies and meaningful embeddings, effectively solving cooperative tasks.

DriveAgent: Multi-Agent Structured Reasoning with LLM and Multimodal Sensor Fusion for Autonomous Driving Xinmeng Hou 2,, Wuqi Wang 1,, Long Y ang 1, Hao Lin 3, Jinglun Feng 4,, Haigen Min 1,, Xiangmo Zhao 1 Abstract -- We introduce DriveAgent, a novel multi-agent autonomous driving framework that leverages large language model (LLM) reasoning combined with multimodal sensor fusion to enhance situational understanding and decision-making. DriveAgent uniquely integrates diverse sensor modalities--including camera, LiDAR, GPS, and IMU--with LLMdriven analytical processes structured across specialized agents. The framework operates through a modular agent-based pipeline comprising four principal modules: (i) a descriptive analysis agent identifying critical sensor data events based on filtered timestamps, (ii) dedicated vehicle-level analysis conducted by LiDAR and vision agents that collaboratively assess vehicle conditions and movements, (iii) environmental reasoning and causal analysis agents explaining contextual changes and their underlying mechanisms, and (iv) an urgency-aware decision-generation agent prioritizing insights and proposing timely maneuvers. This modular design empowers the LLM to effectively coordinate specialized perception and reasoning agents, delivering cohesive, interpretable insights into complex autonomous driving scenarios. Extensive experiments on challenging autonomous driving datasets demonstrate that DriveAgent is achieving superior performance on multiple metrics against baseline methods. These results validate the efficacy of the proposed LLM-driven multi-agent sensor fusion framework, underscoring its potential to substantially enhance the robustness and reliability of autonomous driving systems. 1 I. I NTRODUCTION Promising progress has been made in autonomous driving (AD) in recent years; however, some challenging problems in AD have yet to be solved, especially under dynamic, mul-timodal environments, such as contextual understanding and interpretability [1]. Commonly adopted AD architectures, whether modular or end-to-end, often struggle to integrate insights across heterogeneous sensor modalities--such as cameras, LiDAR, IMU and GPS--especially in edge cases where visual information is ambiguous or missing [2]. 1 Wuqi Wang, Long Y ang, Haigen Min, and Xiangmo Zhao are with Chang'an University, Xi'an, Shaanxi, China. 2 Xinmeng Hou is with Chang'an University, Xi'an, Shaanxi, China and Agency for Science, Technology and Research (A*ST AR), Singapore.

Recently, large language model based (LLM-based) agents have been widely applied across various fields. As a critical part, their memory capabilities have captured significant interest from both industrial and academic communities. Despite the proposal of many advanced memory models in recent research, however, there remains a lack of unified implementations under a general framework. To address this issue, we develop a unified and modular library for developing advanced memory models of LLM-based agents, called MemEngine. Based on our framework, we implement abundant memory models from recent research works. Additionally, our library facilitates convenient and extensible memory development, and offers user-friendly and pluggable memory usage. For benefiting our community, we have made our project publicly available at https://github.com/nuster1128/MemEngine.

Free-form protocols are essential for AI's task generalization but enable new threats like secret collusion and coordinated swarm attacks. Network effects can rapidly spread privacy breaches, disinformation, jailbreaks, and data poisoning, while multi-agent dispersion and stealth optimization help adversaries evade oversight--creating novel persistent threats at a systemic level. Despite their critical importance, these security challenges remain understudied, with research fragmented across disparate fields including AI security, multi-agent learning, complex systems, cybersecurity, game theory, distributed systems, and technical AI governance. We introduce multi-agent security, a new field dedicated to securing networks of decentralized AI agents against threats that emerge or amplify through their interactions--whether direct or indirect via shared environments--with each other, humans, and institutions, and characterise fundamental security-performance trade-offs. Our preliminary work (1) taxonomizes the threat landscape arising from interacting AI agents, (2) surveys security-performance tradeoffs in decentralized AI systems, and (3) proposes a unified research agenda addressing open challenges in designing secure agent systems and interaction environments. By identifying these gaps, we aim to guide research in this critical area to unlock the socioeconomic potential of large-scale agent deployment on the internet, foster public trust, and mitigate national security risks in critical infrastructure and defense contexts.Figure 1: Multi-agent threats demand multi-agent security: [Left] Two malicious AI agents (Mallory and Trudy) are interacting with a human user (Bob) through a shared message board seemingly innocuously to the overseer (magnifying glass).

This paper presents Tunnel, a simple, open source, reinforcement learning training environment for high performance aircraft. It integrates the F16 3D nonlinear flight dynamics into OpenAI Gymnasium python package. The template includes primitives for boundaries, targets, adversaries and sensing capabilities that may vary depending on operational need. This offers mission planners a means to rapidly respond to evolving environments, sensor capabilities and adversaries for autonomous air combat aircraft. It offers researchers access to operationally relevant aircraft physics. Tunnel code base is accessible to anyone familiar with Gymnasium and/or those with basic python skills. This paper includes a demonstration of a week long trade study that investigated a variety of training methods, observation spaces, and threat presentations. This enables increased collaboration between researchers and mission planners which can translate to a national military advantage. As warfare becomes increasingly reliant upon automation, software agility will correlate with decision advantages. Airmen must have tools to adapt to adversaries in this context. It may take months for researchers to develop skills to customize observation, actions, tasks and training methodologies in air combat simulators. In Tunnel, this can be done in a matter of days.