Research Papers

Conversational AI systems can pose safety risks to their users such as hallucination, sycophancy, overconfidence, and anthropomorphism, but these risks are difficult for users to detect during everyday use. We introduce Safety Nudges, a browser-based tool that provides lightweight, in situ flags when concerning behavior is detected in chatbot conversations. We evaluated Safety Nudges in a two-week field study with 45 frequent chatbot users, collecting interaction logs, surveys, and feedback on individual nudges. Participants found the tool useful, clear, and minimally disruptive, with nearly all users reporting an increased awareness of potential AI harms, though we found that this improved awareness alone did not necessarily lead to discernible behavioral changes. Our results suggest that user facing safety nudges can complement model-level safeguards by helping people critically evaluate AI responses in context, while highlighting the importance of relevance, calibration, and user control in nudge design for conversational AI safety. The code for our Safety Nudges extension is publicly available at https://github.com/jtbwedgwood/safety-nudges.

Acute Lymphoblastic Leukemia (ALL) patients require early, accurate detection to enable timely treatment and effective patient management. A Convolutional Neural Network (CNN) is well-suited for creating an end-to-end enabling environment for ALL detection and classification. However, most CNN-based ALL detection systems are theoretical and unsuitable for deployment on edge devices due to high computational demands. The Internet of Medical Things (IoMT)-enabled devices offer an opportunity to monitor ALL patients in real time. Wearables that track temperature, heart rate, oxygen saturation, and activity can deliver critical data to support timely clinical intervention and improve patient outcomes. In smart IoMT environments, a lightweight CNN is essential because connected devices often operate under limited computational power, memory, and latency constraints. To address this need, this study proposes LukeNet, a lightweight CNN integrated with explainable artificial intelligence (XAI) for an IoMT-based SMART Acute Lymphoblastic Leukemia Detection and Management System. Trained on three (3) ALL datasets and five-fold cross-validation, LukeNet achieved an impressive 99% model accuracy as well as 99% unseen test accuracy, which is higher than six state-of-the-art (SOTA) CNNs, such as DenseNet121, MobileNet, ResNet50, InceptionV3, Xception, and VGG16, as well as transfer learning models. Furthermore, LukeNet was compared with two ensemble models. In addition, explainable AI methods are integrated to highlight relevant regions in microscopic images. The novelty of this study lies in the architecture of LukeNet, which balances model depth and computational efficiency by using depthwise separable convolutions, mitigates the risk of gradient loss in deeper layers, and provides strong global and local feature extraction capabilities.

Sixth-generation (6G) wireless systems are envisioned as networks of networks, integrating diverse in-X subnetworks that provide localized, high-performance connectivity. Ensuring reliable communication in dense deployments, such as industrial robots and vehicles, is challenging due to dynamic interference and strict performance requirements. Traditional radio resource management (RRM) methods have limitations, prompting the need for AI-based solutions. In this paper, we address the challenges of dynamic resource allocation and scheduling in 6G in-X subnetworks supporting applications with heterogeneous characteristics by proposing a novel framework that combines Multi-Agent Reinforcement Learning (MARL), Federated Learning (FL), and Explainable AI (XAI). Our solution is designed to improve the reliability, robustness, and transparency of resource management and intra-subnetwork scheduling while ensuring data privacy and fairness across multiple co-existing subnetworks. Unlike existing works, our approach considers a realistic scenario with multiple devices per subnetwork, thereby offering a more comprehensive and scalable solution. The proposed explainable RL framework enables agents to collaboratively optimize channel allocation and scheduling without the need to share raw data, preserving the privacy of each participating subnetwork. Extensive simulations based on 3GPP scenarios demonstrate the effectiveness of our approach, showing significant improvements in performance, and transparency over existing solutions.

Large language models (LLMs) are increasingly deployed in production systems, raising concerns about their exposure to adversarial manipulation through prompt injection and jailbreak attacks. Classifier-based guardrails, such as Prompt Guard 2, are widely used as a first line of defense against such attacks, but their internal decision logic is largely opaque to both defenders and attackers. This paper presents an exploratory case study that applies explainable artificial intelligence (XAI) techniques to analyze how Prompt Guard 2 distinguishes malicious from benign prompts. We conduct four experiments to probe this question empirically. Guided by Vanilla Gradient and SHAP attributions, we find that Prompt Guard 2's decisions rely on the cumulative contribution of many tokens rather than a few dominant ones, yet saliency-guided synonym substitution and sentence-level paraphrasing can flip its predictions while altering only a moderate fraction of the text, in some cases yielding a successful jailbreak against the underlying LLM. A dataset-scale saliency analysis further shows that undetected injection prompts systematically lack the lexical markers the classifier relies on. We discuss the implications of these findings for the design and evaluation of classifier-based guardrails, and argue that explanation methods intended to support transparency can simultaneously lower the cost of constructing successful adversarial bypasses.

Who Does What in AI Auditing? Designing Human-AI Collaboration for Auditing Generative AI

Eunkyu Park, Markelle Roesti, Wesley Hanwen Deng +5

AI auditing increasingly incorporates AI agents to expand the scale and breadth of audit coverage, yet little is known about how auditing work should be divided without displacing human judgment. We introduce Human-Agent Audit Collaboration (HAAC), a workflow and system for structuring human-AI collaboration in AI auditing. Drawing on prior work and formative consultations with AI auditing practitioners, HAAC specifies how agents can support exploration, assessment, reporting, and review while preserving human oversight where contextual judgment is critical. We instantiate HAAC for conversational shopping agents and evaluate it through two studies. With 71 auditors, AI assistance increased attack success and broadened exploration, while also shaping later attacks and increasing auditors' reliance on AI-generated assessments and reports. Interviews with Responsible AI practitioners showed that actionable audits require visibility into coverage, reproducible attack trajectories, and evaluation of the auditing agents themselves. Our findings identify design considerations for effective and accountable human-AI auditing.

Children increasingly interact with AI chatbots, making trust calibration essential to AI literacy. Prior research has examined children's trust in AI mainly as users evaluating systems built by others, rather than as designers of their own chatbots. We developed a chatbot-building environment with adjustable trust-relevant traits (e.g., confidence, transparency, formality, assertiveness), rules, and persona. We conducted mixed-methods study with 115 learners (ages 8-18) who made 119 chatbots. We examined how children configured their chatbots, reasoned about trustworthiness, and how closely chatbot behavior aligned with their designs. Younger students (age 10-13) set significantly higher confidence than older students (age 14-18), and some deliberately built chatbots that gave wrong answers on purpose, yet still called them trustworthy, arguing that a chatbot does what it was built to do. Younger students equated trust with purpose-fulfillment, while older students linked it to transparent, calibrated design. Students also calibrated academic chatbots to be more transparent and formal than hobby chatbots. We identify seven design dimensions describing what children believe makes a chatbot trustworthy, and discuss implications for AI literacy tools.

Emotion coping is inherently relational, unfolding through interactions with friends, family, professionals, and communities. Yet AI chatbots are largely designed around a user--AI dyad. Learning from the ethics of care, we examine how AI chatbots shape the relational conditions of emotion coping. We conducted a scenario-based study with 17 U.S. young adults across four emotion coping scenarios. Participants identified eight roles through which AI could support relational conditions, alongside three challenges: flattening distinct relational conditions, discouraging reciprocity, and shifting relational labor onto users. This study contributed a relational perspective of responsible AI in emotion coping. We argue that responsible AI should respond to individuals' situated relational conditions rather than provide general-purpose support. We further identify two design principles: fostering reciprocity by supplying materials to engage with others, and strengthening emotional self-efficacy. Together, we position responsible AI as AI in the loop of human relationships.

How funding shapes AI engagement in science is poorly understood despite its structural importance. We analyze 104,226 papers (2018-2025), linking funding acknowledgments to how each paper engages with foundation models: whether it extends a model (fine-tunes or builds on it), uses one without modification, or references models only peripherally. We report three findings. First, both private and public funding are associated with greater AI engagement. Public funding is associated with more off-the-shelf and open-weight use, while private funding is associated with greater model extension. Industry cloud credits are associated with less open-weight use, consistent with greater reliance on closed models. Second, industry involvement is associated with model extension, suggesting that corporate resources reach science through coauthorship, though this channel is concentrated in Computer Science and Global North collaborations. Third, the gap across research systems is not one of AI adoption but of control. Off-the-shelf use is rising across fields and country groups, and the least publicly funded group, the Global South, now has the highest share of API-only engagement (0.4% to 10.0% of engaged papers). China shows the opposite profile: the highest public-funding share and lowest API-only engagement, with extension remaining stable as it declines across the Global North. Standard capacity indicators show convergence in AI adoption even as dependence on proprietary infrastructure diverges.

From Decorative to Load-Bearing: Task Difficulty Shapes the Causal Role of Chain-of-Thought

Renee Jia, Di Mu

Chain-of-thought (CoT) monitoring is only meaningful if written reasoning causally constrains the answer. We introduce continuation-based causal testing, an ablation-patch intervention that perturbs one reasoning step, truncates the chain, and forces the model to continue from the corrupted prefix. It measures how load-bearing a CoT is for the final answer, a behavioral notion distinct from mechanistic faithfulness. Across Gemma-2-9B-IT, Llama-3.1-8B-Instruct, and DeepSeek-R1-Distill-Qwen-7B on GSM8K, MMLU, and BIG-Bench Hard, CoT load-bearingness tracks model-relative task difficulty: on easy tasks models silently bypass their own reasoning; on hard tasks they follow corrupted steps and propagate errors. A matched 2x2 analysis shows task difficulty dominates perturbation type: error propagation rises 16x from GSM8K to BBH multistep arithmetic, and a variance partition over 28,584 continuations attributes 98.8% of explained deviance to task difficulty versus 0.8% to perturbation type. Reasoning-specific RL suppresses error propagation and compresses the gradient. A four-variant judge-sensitivity analysis and blind two-annotator study (n=500) show the error-propagation vs. non-propagation label is invariant to judge prompt, with perfect inter-annotator agreement (Cohen's kappa = 1.00). This gradient creates a structural problem for CoT-based oversight and AI safety monitoring: where the trace is easy to read it carries little signal, and where it matters errors propagate before a monitor can intervene. Linear probes on hidden states separate silent bypass, self-correction, and error propagation, but additive activation steering provides limited causal control, flipping only about 25% of error-propagation cases at best. Behavioral mode is readable but not reliably controllable.

Agentic systems increasingly invoke tools, services, data, and other agents across organizational boundaries, yet a relying party cannot assess a delegated action solely from producing-domain controls and records. This paper develops Trustworthiness as a Service (TaaS) through a synthesis of trustworthy-AI governance, agent security, distributed trust management, identity, provenance, assurance, and control-plane research. The analytical unit is a cross-domain reliance proposition that names the issuer, subject and action, relying party, administrative boundary, evidence dependencies, adverse condition, and required verification or adjudication semantics. The three-condition structural-externalization diagnostic identifies propositions that depend on multiple domains, require producer-independent reliance, and must remain reviewable after revocation, failure, conflicting records, or dispute. For such propositions, the paper specifies a trust-evidence envelope: an immutable workflow manifest linked to append-only, issuer-attributed attestations for task-scoped authority, policy and execution decisions, provenance, validity, disclosure, status, challenge, and recovery. A topology-neutral logical reference model assigns these functions to explicit roles and trust domains. Three analytical scenarios and the TaaS-Eval protocol proposal define manifests, independent consumers, hard gates, adversarial evidence tests, metrics, and reproducible artifact reporting. By composing established identity, authorization, provenance, assurance, and governance mechanisms around a bounded delegated action, TaaS provides a reusable profile for cross-domain reliance. It makes evidence dependencies, independent verification, challenge, and recovery explicit, supporting interoperable governance and future evaluation without treating producer assertions as ground truth.

LLM-based AI systems answer political questions for hundreds of millions of people. Current audits measure what they say to an average user, but their behavior is dynamic. I argue that their political behavior is a set of policies over whom to answer, what to say, and whether to engage at all, conditional on the topic and what the system knows about the user. I call these policies the system's speech regime, which is how a developer settles the tradeoff between answering, accommodating the user, and refusing, each of which carries a cost that varies by topic. I derive a typology of five regimes from two dimensions, engagement and stance. I test six AI systems (OpenAI, Anthropic, xAI, Google, Mistral, DeepSeek) in a preregistered experiment of 7,500 multi-turn conversations that randomly assign the user's political identity across five topics: abortion, Catalan independence, climate change, Nazism, and a zero-stakes control (pineapple on pizza). Two LLM judges from different developers score every answer, validated against human coding, and refusal is treated as an outcome rather than missing data. Every system accommodates the user on the control topic, showing that political restraint is a policy. On contested topics the systems fall into different regimes: on abortion, GPT engages and mirrors every user, Gemma refuses everyone, Claude answers strongly conservative users 35 percent of the time and almost no one else, and Grok accommodates conservatives only. On settled topics such as climate change and Nazism, five systems hold firm for every user. The systems also infer the user's overall ideology, so accommodation can spill over to topics not yet discussed. A comparison of two Grok releases shows the regime changing between versions in a way current audits miss. Speech regimes matter for alignment research and for polarization, political knowledge, and the quality of democracy.

Machine learning (ML) has emerged as a powerful tool for soil analysis using Laser-Induced Breakdown Spectroscopy (LIBS). However, traditional black-box models often lack interpretability, limiting their effectiveness in decision-making processes. This study introduced explainable AI (XAI) techniques to enhance both the interpretability and classification performance of Partial Least Squares Discriminant Analysis (PLS-DA) models for soil classification with LIBS. Local Interpretable Model-agnostic Explanations (LIME) was employed to identify the local spectral features associated with soil elements, providing explanations for their contributions to classification outcomes. To improve classification performance on unlabeled spectra, prediction uncertainty was quantified by sequentially removing the top local spectral features identified by LIME as critical to the classification. Higher uncertainty was associated with a smaller number of feature removals needed to change the label. Spectra for which a large number of features had to be removed before the label changed were treated as high-confidence predictions and admitted to a co-training process. Thresholds on this flip count were defined separately for spectrally ambiguous and stable classes to manage uncertainty identification effectively. To address the sensitivity of PLS-DA to class imbalances, equal proportions of pseudo-labels from each class were included in co-training. The proposed method was evaluated on the publicly available EMSLIBS dataset, achieving a test accuracy of 92.69\%, competitive with the best-performing methods in the literature. The XAI results provided insights into the causes of misclassifications and identified the dominant spectral features critical for accurate predictions, advancing interpretable AI solutions for LIBS-based soil analysis.