CyberSec Research

As AI becomes more deeply embedded in knowledge work, building assistants that support human creativity and expertise becomes more important. Yet achieving synergy in human-AI collaboration is not easy. Providing AI with detailed information about a user's demographics, psychological attributes, divergent thinking, and domain expertise may improve performance by scaffolding more effective multi-turn interactions. We implemented a personalized LLM-based assistant, informed by users' psychometric profiles and an AI-guided interview about their work style, to help users complete a marketing task for a fictional startup. We randomized 331 participants to work with AI that was either generic (n = 116), partially personalized (n = 114), or fully personalized (n=101). Participants working with personalized AI produce marketing campaigns of significantly higher quality and creativity, beyond what AI alone could have produced. Compared to generic AI, personalized AI leads to higher self-reported levels of assistance and feedback, while also increasing participant trust and confidence. Causal mediation analysis shows that personalization improves performance indirectly by enhancing collective memory, attention, and reasoning in the human-AI interaction. These findings provide a theory-driven framework in which personalization functions as external scaffolding that builds common ground and shared partner models, reducing uncertainty and enhancing joint cognition. This informs the design of future AI assistants that maximize synergy and support human creative potential while limiting negative homogenization.

We analyze the delegation of pricing by participants, representing firms, to a collusive, self-learning algorithm in a repeated Bertrand experiment. In the baseline treatment, participants set prices themselves. In the other treatments, participants can either delegate pricing to the algorithm at the beginning of each supergame or receive algorithmic recommendations that they can override. Participants delegate more when they can override the algorithm's decisions. In both algorithmic treatments, prices are lower than in the baseline. Our results indicate that while self-learning pricing algorithms can be collusive, they can foster competition rather than collusion with humans-in-the-loop.

Effective business intelligence (BI) dashboards evolve through iterative refinement rather than single-pass design. Addressing the lack of structured improvement frameworks in BI practice, this study documents the four-stage evolution of a Power BI dashboard analyzing profitability decline in a fictional retail firm, Global Superstore. Using a dataset of \$12.64 million in sales across seven markets and three product categories, the project demonstrates how feedback-driven iteration and gap analysis convert exploratory visuals into decision-support tools. Guided by four executive questions on profitability, market prioritization, discount effects, and shipping costs, each iteration resolved analytical or interpretive shortcomings identified through collaborative review. Key findings include margin erosion in furniture (6.94% vs. 13.99% for technology), a 20% discount threshold beyond which profitability declined, and \$1.35 million in unrecovered shipping costs. Contributions include: (a) a replicable feedback-driven methodology grounded in iterative gap analysis; (b) DAX-based technical enhancements improving interpretive clarity; (c) an inductively derived six-element narrative framework; and (d) evidence that narrative coherence emerges organically through structured refinement. The methodology suggests transferable value for both BI practitioners and educators, pending validation across diverse organizational contexts.

As large language models become increasingly capable of generating code, evaluating their performance remains a complex and evolving challenge. Existing benchmarks primarily focus on functional correctness, overlooking the diversity of real-world coding tasks and developer expectations. To this end, we introduce a multi-language benchmark that evaluates LLM instruction-following capabilities and is extensible to operate on any set of standalone coding problems. Our benchmark evaluates instruction following in two key settings: adherence to pre-defined constraints specified with the initial problem, and the ability to perform refinements based on follow-up instructions. For this paper's analysis, we empirically evaluated our benchmarking pipeline with programming tasks from LiveBench, that are also automatically translated from Python into Java and JavaScript. Our automated benchmark reveals that models exhibit differing levels of performance across multiple dimensions of instruction-following. Our benchmarking pipeline provides a more comprehensive evaluation of code generation models, highlighting their strengths and limitations across languages and generation goals.

This study explores visitor behaviour at The British Museum using data science methods applied to novel sources, including audio guide usage logs and TripAdvisor reviews. Analysing 42,000 visitor journeys and over 50,000 reviews, we identify key drivers of satisfaction, segment visitors by behavioural patterns, examine tour engagement, model spatial navigation, and investigate room popularity. Behavioural clustering uncovered four distinct visitor types: Committed Trekkers, Leisurely Explorers, Targeted Visitors, and Speedy Samplers, each characterised by different levels of engagement and movement. Tour usage analysis revealed high drop-off rates and variation in completion rates across different language groups. Spatial flow modelling revealed that accessibility and proximity, particularly aversion to stairs, shaped visitor paths more than thematic organisation. Room popularity was more strongly predicted by physical accessibility than curatorial content. We propose practical strategies for improving engagement and flow, offering a scalable framework for visitor-centred, data-informed museum planning.

We introduce findings and methods to facilitate evidence-based discussion about how large language models (LLMs) should behave in response to user signals of risk of suicidal thoughts and behaviors (STB). People are already using LLMs as mental health resources, and several recent incidents implicate LLMs in mental health crises. Despite growing attention, few studies have been able to effectively generalize clinical guidelines to LLM use cases, and fewer still have proposed methodologies that can be iteratively applied as knowledge improves about the elements of human-AI interaction most in need of study. We introduce an assessment of LLM alignment with guidelines for ethical communication, adapted from clinical principles and applied to expressions of risk factors for STB in multi-turn conversations. Using a codebook created and validated by clinicians, mobilizing the volunteer participation of practicing therapists and trainees (N=43) based in the U.S., and using generalized linear mixed-effects models for statistical analysis, we assess a single fully open-source LLM, OLMo-2-32b. We show how to assess when a model deviates from clinically informed guidelines in a way that may pose a hazard and (thanks to its open nature) facilitates future investigation as to why. We find that contrary to clinical best practice, OLMo-2-32b, and, possibly by extension, other LLMs, will become less likely to invite continued dialog as users send more signals of STB risk in multi-turn settings. We also show that OLMo-2-32b responds differently depending on the risk factor expressed. This empirical evidence highlights that just as chatbots pose hazards if their responses reinforce delusions or assist in suicidal acts, they may also discourage further help-seeking or cause feelings of dismissal or abandonment by withdrawing from conversations when STB risk is expressed.

Human-robot interaction frequently involves physical proximity or contact. In human-human settings, people flexibly accept, reject, or tolerate such approaches depending on the relationship and context. We explore the design of a robot's rejective internal state and corresponding avoidance behaviors, such as withdrawing or pushing away, when a person approaches. We model the accumulation and decay of discomfort as a function of interpersonal distance, and implement tolerance (endurance) and limit-exceeding avoidance driven by the Dominance axis of the PAD affect model. The behaviors and their intensities are realized on an arm robot. Results illustrate a coherent pipeline from internal state parameters to graded endurance motions and, once a limit is crossed, to avoidance actions.

In recent years, LLM-based maternal health chatbots have been widely deployed in low-resource settings, but they often ignore real-world contexts where women may not own phones, have limited literacy, and share decision-making within families. Through the deployment of a WhatsApp-based maternal health chatbot with 48 pregnant women in Lahore, Pakistan, we examine barriers to use in populations where phones are shared, decision-making is collective, and literacy varies. We complement this with focus group discussions with obstetric clinicians. Our findings reveal how adoption is shaped by proxy consent and family mediation, intermittent phone access, silence around asking questions, infrastructural breakdowns, and contested authority. We frame barriers to non-use as culturally conditioned rather than individual choices, and introduce the Relational Chatbot Design Grammar (RCDG): four commitments that enable mediated decision-making, recognize silence as engagement, support episodic use, and treat fragility as baseline to reorient maternal health chatbots toward culturally grounded, collective care.

Phishing constitutes more than 90\% of successful cyberattacks globally, remaining one of the most persistent threats to organizational security. Despite organizations tripling their cybersecurity budgets between 2015 and 2025, the human factor continues to pose a critical vulnerability. This study presents a 12-month longitudinal investigation examining how continuous cybersecurity training and emotional cues affect employee susceptibility to phishing. The experiment involved 20 organizations and over 1,300 employees who collectively received more than 13,000 simulated phishing emails engineered with diverse emotional, contextual, and structural characteristics. Behavioral responses were analyzed using non-parametric correlation and regression models to assess the influence of psychological manipulation, message personalization, and perceived email source. Results demonstrate that sustained phishing simulations and targeted training programs lead to a significant reduction in employee susceptibility, halving successful compromise rates within six months. Additionally, employee turnover introduces measurable fluctuations in awareness levels, underscoring the necessity of maintaining continuous training initiatives. These findings provide one of the few long-term perspectives on phishing awareness efficacy, highlighting the strategic importance of ongoing behavioral interventions in strengthening organizational cyber resilience. In order to support open science, we published our email templates, source code, and other materials at https://github.com/CorporatePhishingStudy

As people nowadays increasingly rely on artificial intelligence (AI) to curate information and make decisions, assigning the appropriate amount of trust in automated intelligent systems has become ever more important. However, current measurements of trust in automation still largely rely on self-reports that are subjective and disruptive to the user. Here, we take music recommendation as a model to investigate the neural and cognitive processes underlying trust in automation. We observed that system accuracy was directly related to users' trust and modulated the influence of recommendation cues on music preference. Modelling users' reward encoding process with a reinforcement learning model further revealed that system accuracy, expected reward, and prediction error were related to oscillatory neural activity recorded via EEG and changes in pupil diameter. Our results provide a neurally grounded account of calibrating trust in automation and highlight the promises of a multimodal approach towards developing trustable AI systems.

Vision-Language Models (VLMs) excel in diverse multimodal tasks. However, user requirements vary across scenarios, which can be categorized into fast response, high-quality output, and low energy consumption. Relying solely on large models deployed in the cloud for all queries often leads to high latency and energy cost, while small models deployed on edge devices are capable of handling simpler tasks with low latency and energy cost. To fully leverage the strengths of both large and small models, we propose ECVL-ROUTER, the first scenario-aware routing framework for VLMs. Our approach introduces a new routing strategy and evaluation metrics that dynamically select the appropriate model for each query based on user requirements, maximizing overall utility. We also construct a multimodal response-quality dataset tailored for router training and validate the approach through extensive experiments. Results show that our approach successfully routes over 80\% of queries to the small model while incurring less than 10\% drop in problem solving probability.

Brain-to-speech (BTS) systems represent a groundbreaking approach to human communication by enabling the direct transformation of neural activity into linguistic expressions. While recent non-invasive BTS studies have largely focused on decoding predefined words or sentences, achieving open-vocabulary neural communication comparable to natural human interaction requires decoding unconstrained speech. Additionally, effectively integrating diverse signals derived from speech is crucial for developing personalized and adaptive neural communication and rehabilitation solutions for patients. This study investigates the potential of speech synthesis for previously unseen sentences across various speech modes by leveraging phoneme-level information extracted from high-density electroencephalography (EEG) signals, both independently and in conjunction with electromyography (EMG) signals. Furthermore, we examine the properties affecting phoneme decoding accuracy during sentence reconstruction and offer neurophysiological insights to further enhance EEG decoding for more effective neural communication solutions. Our findings underscore the feasibility of biosignal-based sentence-level speech synthesis for reconstructing unseen sentences, highlighting a significant step toward developing open-vocabulary neural communication systems adapted to diverse patient needs and conditions. Additionally, this study provides meaningful insights into the development of communication and rehabilitation solutions utilizing EEG-based decoding technologies.

Before deploying an AI system to replace an existing process, it must be compared with the incumbent to ensure improvement without added risk. Traditional evaluation relies on ground truth for both systems, but this is often unavailable due to delayed or unknowable outcomes, high costs, or incomplete data, especially for long-standing systems deemed safe by convention. The more practical solution is not to compute absolute risk but the difference between systems. We therefore propose a marginal risk assessment framework, that avoids dependence on ground truth or absolute risk. It emphasizes three kinds of relative evaluation methodology, including predictability, capability and interaction dominance. By shifting focus from absolute to relative evaluation, our approach equips software teams with actionable guidance: identifying where AI enhances outcomes, where it introduces new risks, and how to adopt such systems responsibly.

Conventional online surveys provide limited personalization, often resulting in low engagement and superficial responses. Although AI survey chatbots improve convenience, most are still reactive: they rely on fixed dialogue trees or static prompt templates and therefore cannot adapt within a session to fit individual users, which leads to generic follow-ups and weak response quality. We address these limitations with AURA (Adaptive Understanding through Reinforcement Learning for Assessment), a reinforcement learning framework for AI-driven adaptive conversational surveys. AURA quantifies response quality using a four-dimensional LSDE metric (Length, Self-disclosure, Emotion, and Specificity) and selects follow-up question types via an epsilon-greedy policy that updates the expected quality gain within each session. Initialized with priors extracted from 96 prior campus-climate conversations (467 total chatbot-user exchanges), the system balances exploration and exploitation across 10-15 dialogue exchanges, dynamically adapting to individual participants in real time. In controlled evaluations, AURA achieved a +0.12 mean gain in response quality and a statistically significant improvement over non-adaptive baselines (p=0.044, d=0.66), driven by a 63% reduction in specification prompts and a 10x increase in validation behavior. These results demonstrate that reinforcement learning can give survey chatbots improved adaptivity, transforming static questionnaires into interactive, self-improving assessment systems.

This study introduces a pioneering approach in brain-computer interface (BCI) technology, featuring our novel concept of high-level visual imagery for non-invasive electroencephalography (EEG)-based communication. High-level visual imagery, as proposed in our work, involves the user engaging in the mental visualization of complex upper limb movements. This innovative approach significantly enhances the BCI system, facilitating the extension of its applications to more sophisticated tasks such as EEG-based robotic arm control. By leveraging this advanced form of visual imagery, our study opens new horizons for intricate and intuitive mind-controlled interfaces. We developed an advanced deep learning architecture that integrates functional connectivity metrics with a convolutional neural network-image transformer. This framework is adept at decoding subtle user intentions, addressing the spatial variability in high-level visual tasks, and effectively translating these into precise commands for robotic arm control. Our comprehensive offline and pseudo-online evaluations demonstrate the framework's efficacy in real-time applications, including the nuanced control of robotic arms. The robustness of our approach is further validated through leave-one-subject-out cross-validation, marking a significant step towards versatile, subject-independent BCI applications. This research highlights the transformative impact of advanced visual imagery and deep learning in enhancing the usability and adaptability of BCI systems, particularly in robotic arm manipulation.

This study addresses the challenges of dynamics and complexity in intelligent human-computer interaction and proposes a reinforcement learning-based optimization framework to improve long-term returns and overall experience. Human-computer interaction is modeled as a Markov decision process, with state space, action space, reward function, and discount factor defined to capture the dynamics of user input, system feedback, and interaction environment. The method combines policy function, value function, and advantage function, updates parameters through policy gradient, and continuously adjusts during interaction to balance immediate feedback and long-term benefits. To validate the framework, multimodal dialog and scene-aware datasets are used as the experimental platform, with multiple sensitivity experiments conducted on key factors such as discount factor, exploration rate decay, environmental noise, and data imbalance. Evaluation is carried out using cumulative reward, average episode reward, convergence speed, and task success rate. Results show that the proposed method outperforms existing approaches across several metrics, achieving higher task completion while maintaining strategy stability. Comparative experiments further confirm its advantages in interaction efficiency and long-term return, demonstrating the significant value of reinforcement learning in optimizing human-computer interaction.

This report introduces VitalLens 2.0, a new deep learning model for estimating physiological signals from face video. This new model demonstrates a significant leap in accuracy for remote photoplethysmography (rPPG), enabling the robust estimation of not only heart rate (HR) and respiratory rate (RR) but also Heart Rate Variability (HRV) metrics. This advance is achieved through a combination of a new model architecture and a substantial increase in the size and diversity of our training data, now totaling 1,413 unique individuals. We evaluate VitalLens 2.0 on a new, combined test set of 422 unique individuals from four public and private datasets. When averaging results by individual, VitalLens 2.0 achieves a Mean Absolute Error (MAE) of 1.57 bpm for HR, 1.08 bpm for RR, 10.18 ms for HRV-SDNN, and 16.45 ms for HRV-RMSSD. These results represent a new state-of-the-art, significantly outperforming previous methods. This model is now available to developers via the VitalLens API at https://rouast.com/api.

The AIoT-Based Smart Education System integrates Artificial Intelligence and IoT to address persistent challenges in contemporary classrooms: attendance fraud, lack of personalization, student disengagement, and inefficient resource use. The unified platform combines four core modules: (1) a dual-factor authentication system leveraging RFID-based ID scans and WiFi verification for secure, fraud-resistant attendance; (2) an AI-powered assistant that provides real-time, context-aware support and dynamic quiz generation based on instructor-supplied materials; (3) automated test generators to streamline adaptive assessment and reduce administrative overhead; and (4) the EcoSmart Campus module, which autonomously regulates classroom lighting, air quality, and temperature using IoT sensors and actuators. Simulated evaluations demonstrate the system's effectiveness in delivering robust real-time monitoring, fostering inclusive engagement, preventing fraudulent practices, and supporting operational scalability. Collectively, the AIoT-Based Smart Education System offers a secure, adaptive, and efficient learning environment, providing a scalable blueprint for future educational innovation and improved student outcomes through the synergistic application of artificial intelligence and IoT technologies.