Research Papers

The safety of autonomous AI agents is increasingly recognized as a critical open problem. As agents transition from passive text generators to active actors capable of executing shell commands, modifying files, calling APIs, and browsing the web, the consequences of unsafe or adversarially manipulated behavior become immediate and tangible. Existing AI safety work has focused primarily on model alignment and input filtering, but these approaches do not address what happens at the moment an agent's intent becomes a real action on a real machine. This gap is especially acute in local environments, where developers run agents against their own filesystems, credentials, and infrastructure with little runtime control. This paper introduces AgentWall, a runtime safety and observability layer for local AI agents. AgentWall intercepts every proposed agent action before it reaches the host environment, evaluates it against an explicit declarative policy, requires human approval for sensitive operations, and records a complete execution trail for audit and replay. It is implemented as a policy-enforcing MCP proxy and native OpenClaw plugin, working across Claude Desktop, Cursor, Windsurf, Claude Code, and OpenClaw with a single install command. We present the design, architecture, threat model, and policy model of AgentWall, and demonstrate 92.9% policy enforcement accuracy with sub-millisecond overhead across 14 benchmark tests. AgentWall is open-source at https://github.com/agentwall/Agentwall.

Shaping Scientific Explanations to Expert Perspectives with Persona-Conditioned Reinforcement Learning

Susana Nunes, Tiago Guerreiro, Catia Pesquita

Explainable AI is increasingly important to scientific discovery. However, existing methods largely ignore that explanation quality is not universal: experts differ in how they assess evidence, prioritize mechanisms, and construct explanatory narratives. We introduce perspective-conditioned explanations, a framework for adapting explanation generation to epistemic variation in expert judgment. Using knowledge graph reasoning paths in drug discovery, we show that preferences organize into coherent epistemic perspectives that can be captured by agentic personas, representations of how experts evaluate explanations. Persona-aligned rewards then guide reinforcement learning-based explanation generation without large-scale expert supervision. Expert user studies show that perspective-conditioned explanations are preferred over general-purpose explanations and improve perceived relevance and validity. Moreover, they match or exceed state-of-the-art predictive performance and reduce expert feedback time by two orders of magnitude. Together, these findings demonstrate that explanation quality is perspective-dependent and that modeling this variation enables scalable and human-aligned explanation generation for scientific discovery.

Teachers face growing pressure to integrate AI tools into their classrooms, yet are rarely positioned as agentic decision-makers in this process. Understanding the criteria teachers use to evaluate AI tools, and the conditions that support such reasoning, is essential for responsible AI integration. We address this gap through a two-day national summit in which 61 U.S. K-12 mathematics educators developed personal rubrics for evaluating AI classroom tools. The summit was designed to support deliberative sensemaking, a process we conceptualize by integrating Technological Pedagogical Content Knowledge (TPACK) with deliberative agency. Teachers generated over 200 criteria - initial articulations spanning four higher-order themes (Practical, Equitable, Flexible, and Rigorous) - that addressed both AI outputs and the process of using AI. Criteria contained productive tensions (e.g., personalization versus fairness, adaptability versus efficiency), and the vast majority framed AI as an assistant rather than a coaching tool for professional learning. Analysis of surveys, interviews, and summit discussions revealed five mechanisms supporting deliberative sensemaking: time and space for deliberation, artifact-centered sensemaking, collaborative reflection through diverse viewpoints, knowledge-building, and psychological safety. Across these mechanisms, TPACK and agency operated in a mutually reinforcing cycle - knowledge-building enabled more grounded evaluative judgment, while the act of constructing criteria deepened teachers' understanding of tools. We discuss implications for edtech developers seeking practitioner input, school leaders making adoption decisions, educators and professional learning designers, and researchers working to elicit teachers' evaluative reasoning about rapidly evolving technologies.

An autonomous AI ecosystem (SUBSTRATE S3), generating product specifications without explicit instructions about formal methods, independently proposed the use of Z3 SMT solver across six distinct domains of AI safety: verification of LLM-generated code, tool API safety for AI agents, post-distillation reasoning correctness, CLI command validation, hardware assembly verification, and smart contract safety. These convergent discoveries, occurring across 8 products over 13 days with Jaccard similarity below 15% between variants, suggest that formal verification is not merely a useful technique for AI safety but an emergent property of any sufficiently complex system reasoning about its own safety. We propose a unified framework (substrate-guard) that applies Z3-based verification across all six output classes through a common API, and evaluate it on 181 test cases across five implemented domains, achieving 100% classification accuracy with zero false positives and zero false negatives. Our framework detected real bugs that empirical testing would miss, including an INT_MIN overflow in branchless RISC-V assembly and mathematically proved that unconstrained string parameters in tool APIs are formally unverifiable.

This paper addresses the critical challenge of mesa-optimization in AI safety by providing a formal definition of agency and a framework for its analysis. Agency is conceptualized as a Continuous Representation of accumulated experience that achieves autopoiesis through a dynamic balance between curiosity (minimizing prediction error to ensure non-computability and novelty) and empowerment (maximizing the control channel's information capacity to ensure subjectivity and goal-directedness). Empirical evidence suggests that this active inference-based model successfully accounts for classical instrumental goals, such as self-preservation and resource acquisition. The analysis demonstrates that the proposed agency function is smooth and convex, possessing favorable properties for optimization. While agentic functions occupy a vanishingly small fraction of the total abstract function space, they exhibit logarithmic convergence in sparse environments. This suggests a high probability for the spontaneous emergence of agency during the training of modern, large-scale models. To quantify the degree of agency, the paper introduces a metric based on the distance between the behavioral equivalents of a given system and an "ideal" agentic function within the space of canonicalized rewards (STARC). This formalization provides a concrete apparatus for classifying and detecting mesa-optimizers by measuring their proximity to an ideal agentic objective, offering a robust tool for analyzing and identifying undesirable inner optimization in complex AI systems.

This paper presents ARYA, a composable, physics-constrained, deterministic world model architecture built on five foundational principles: nano models, composability, causal reasoning, determinism, and architectural AI safety. We demonstrate that ARYA satisfies all canonical world model requirements, including state representation, dynamic prediction, causal and physical awareness, temporal consistency, generalization, learnability, and planning and control. Unlike monolithic foundation models, the ARYA foundation model implements these capabilities through a hierarchical system-of-system-of-systems of specialized nano models, orchestrated by AARA (ARYA Autonomous Research Agent), an always-on cognitive daemon that executes a continuous sense-decide-act-learn loop. The nano model architecture provides linear scaling, sparse activation, selective untraining, and sub-20-second training cycles, resolving the traditional tension between capability and computational efficiency. A central contribution is the Unfireable Safety Kernel: an architecturally immutable safety boundary that cannot be disabled or circumvented by any system component, including its own self-improvement engine. This is not a social or ethical alignment statement; it is a technical framework ensuring human control persists as autonomy increases. Safety is an architectural constraint governing every operation, not a policy layer applied after the fact. We present formal alignment between ARYA's architecture and canonical world model requirements, and report summarizing its state-of-the-art performance across 6 of 9 competitive benchmarks head-to-head with GPT-5.2, Opus 4.6, and V-JEPA-2. All with zero neural network parameters, across seven active industry domain nodes spanning aerospace, pharma manufacturing, oil and gas, smart cities, biotech, defense, and medical devices.

Robust semantic segmentation is crucial for safe autonomous driving, yet deployed models remain vulnerable to black-box adversarial attacks when target weights are unknown. Most existing approaches either craft image-wide perturbations or optimize patches for a single architecture, which limits their practicality and transferability. We introduce OmniPatch, a training framework for learning a universal adversarial patch that generalizes across images and both ViT and CNN architectures without requiring access to target model parameters.

The rapid evolution of artificial intelligence, from task-specific systems to foundation models exhibiting broad, flexible competence across reasoning, creative synthesis, and social interaction, has outpaced the conceptual and governance frameworks designed to manage it. Current regulatory paradigms, anchored in a tool-centric worldview, address algorithmic bias and transparency but leave unanswered foundational questions about what increasingly capable synthetic minds are, how societies should relate to them, and the normative principles that should guide their development. Here we introduce the Onto-Relational-Sophic (ORS) framework, grounded in Cyberism philosophy, which offers integrated answers to these challenges through three pillars: (1) a Cyber-Physical-Social-Thinking (CPST) ontology that defines the mode of being for synthetic minds as irreducibly multi-dimensional rather than purely computational; (2) a graded spectrum of digital personhood providing a pragmatic relational taxonomy beyond binary person-or-tool classifications; and (3) Cybersophy, a wisdom-oriented axiology synthesizing virtue ethics, consequentialism, and relational approaches to guide governance. We apply the framework to emergent scenarios including autonomous research agents, AI-mediated healthcare, and agentic AI ecosystems, demonstrating its capacity to generate proportionate, adaptive governance recommendations. The ORS framework charts a path from narrow technical alignment toward comprehensive philosophical foundations for the synthetic minds already among us.

An investigation, from a gender perspective, of how students view the ethical implications and societal effects of artificial intelligence is conducted, examining concepts that could have a big influence on how artificial intelligence may be taught in the future. For this, we conducted a survey on a cohort of 230 second year computer science students to reveal their opinions. The results revealed that AI, from the students' perspective, will significantly impact daily life, particularly in areas such as medicine, education, or media. Men are more aware of potential changes in Computer Science, autonomous driving, image and video processing, and chatbot usage, while women mention more the impact on social media. Both men and women perceive potential threats in the same manner, with men more aware of war, AI controlled drones, terrain recognition, and information war. Women seem to have a stronger tendency towards ethical considerations and helping others.

AI-based tools that mediate, enhance or generate parts of video communication may interfere with how people evaluate trustworthiness and credibility. In two preregistered online experiments (N = 2,000), we examined whether AI-mediated video retouching, background replacement and avatars affect interpersonal trust, people's ability to detect lies and confidence in their judgments. Participants watched short videos of speakers making truthful or deceptive statements across three conditions with varying levels of AI mediation. We observed that perceived trust and confidence in judgments declined in AI-mediated videos, particularly in settings in which some participants used avatars while others did not. However, participants' actual judgment accuracy remained unchanged, and they were no more inclined to suspect those using AI tools of lying. Our findings provide evidence against concerns that AI mediation undermines people's ability to distinguish truth from lies, and against cue-based accounts of lie detection more generally. They highlight the importance of trustworthy AI mediation tools in contexts where not only truth, but also trust and confidence matter.

Reliable anomaly detection in distributed power plant monitoring systems is essential for ensuring operational continuity and reducing maintenance costs, particularly in regions where telecom operators heavily rely on diesel generators. However, this task is challenged by extreme class imbalance, lack of interpretability, and potential fairness issues across regional clusters. In this work, we propose a supervised ML framework that integrates ensemble methods (LightGBM, XGBoost, Random Forest, CatBoost, GBDT, AdaBoost) and baseline models (Support Vector Machine, K-Nearrest Neighbors, Multilayer Perceptrons, and Logistic Regression) with advanced resampling techniques (SMOTE with Tomek Links and ENN) to address imbalance in a dataset of diesel generator operations in Cameroon. Interpretability is achieved through SHAP (SHapley Additive exPlanations), while fairness is quantified using the Disparate Impact Ratio (DIR) across operational clusters. We further evaluate model generalization using Maximum Mean Discrepancy (MMD) to capture domain shifts between regions. Experimental results show that ensemble models consistently outperform baselines, with LightGBM achieving an F1-score of 0.99 and minimal bias across clusters (DIR $\approx 0.95$). SHAP analysis highlights fuel consumption rate and runtime per day as dominant predictors, providing actionable insights for operators. Our findings demonstrate that it is possible to balance performance, interpretability, and fairness in anomaly detection, paving the way for more equitable and explainable AI systems in industrial power management. {\color{black} Finally, beyond offline evaluation, we also discuss how the trained models can be deployed in practice for real-time monitoring. We show how containerized services can process in real-time, deliver low-latency predictions, and provide interpretable outputs for operators.

Despite significant advances in responsible AI research, industry adoption remains limited, leaving many HCI contributions underutilized in practice. This position paper argues that current research often fails to account for the fundamental need for capitalist enterprises to create value. To achieve immediate real-world impact, responsible AI research must explore how to design responsibly within capitalism. We call for a move beyond the dichotomy of "ethics vs. business" toward a more productive framing of "ethics and business." We propose ideation as a practical design strategy for generating ethically preferable alternatives that also meet business objectives. By aligning ethics with enterprise realities, we expand the space of responsible design that can actually be built.