Research Papers

The development of Open-Source Software (OSS) is not only a technical challenge, but also a social one due to the diverse mixture of contributors. To this aim, social-coding platforms, such as GitHub, provide the infrastructure needed to host and develop the code, but also the support for enabling the community's collaboration, which is driven by non-coding contributions, such as issues (i.e., change proposals or bug reports) or comments to existing contributions. As with any other social endeavor, this development process faces ethical challenges, which may put at risk the project's sustainability. To foster a productive and positive environment, OSS projects are increasingly deploying codes of conduct, which define rules to ensure a respectful and inclusive participatory environment, with the Contributor Covenant being the main model to follow. However, monitoring and enforcing these codes of conduct is a challenging task, due to the limitations of current approaches. In this paper, we propose an approach to classify the ethical quality of non-coding contributions in OSS projects by relying on Large Language Models (LLM), a promising technology for text classification tasks. We defined a set of ethical metrics based on the Contributor Covenant and developed a classification approach to assess ethical behavior in OSS non-coding contributions, using prompt engineering to guide the model's output.

Ensuring transparency and trust in AI-driven public health and biomedical sciences systems requires more than accurate predictions-it demands explanations that are clear, contextual, and socially accountable. While explainable AI (XAI) has advanced in areas like feature attribution and model interpretability, most methods still lack the structure and adaptability needed for diverse health stakeholders, including clinicians, policymakers, and the general public. We introduce PHAX-a Public Health Argumentation and eXplainability framework-that leverages structured argumentation to generate human-centered explanations for AI outputs. PHAX is a multi-layer architecture combining defeasible reasoning, adaptive natural language techniques, and user modeling to produce context-aware, audience-specific justifications. More specifically, we show how argumentation enhances explainability by supporting AI-driven decision-making, justifying recommendations, and enabling interactive dialogues across user types. We demonstrate the applicability of PHAX through use cases such as medical term simplification, patient-clinician communication, and policy justification. In particular, we show how simplification decisions can be modeled as argument chains and personalized based on user expertise-enhancing both interpretability and trust. By aligning formal reasoning methods with communicative demands, PHAX contributes to a broader vision of transparent, human-centered AI in public health.

Exploring the socio-technical challenges confronting the adoption of AI in organisational settings is something that has so far been largely absent from the related literature. In particular, research into requirements for trustworthy AI typically overlooks how people deal with the problems of trust in the tools that they use as part of their everyday work practices. This article presents some findings from an ongoing ethnographic study of how current tools are used in air traffic control work and what it reveals about requirements for trustworthy AI in air traffic control and other safety-critical application domains.

This paper introduces the TAI Scan Tool, a RAG-based TAI self-assessment tool with minimalistic input. The current version of the tool supports the legal TAI assessment, with a particular emphasis on facilitating compliance with the AI Act. It involves a two-step approach with a pre-screening and an assessment phase. The assessment output of the system includes insight regarding the risk-level of the AI system according to the AI Act, while at the same time retrieving relevant articles to aid with compliance and notify on their obligations. Our qualitative evaluation using use-case scenarios yields promising results, correctly predicting risk levels while retrieving relevant articles across three distinct semantic groups. Furthermore, interpretation of results shows that the tool's reasoning relies on comparison with the setting of high-risk systems, a behaviour attributed to their deployment requiring careful consideration, and therefore frequently presented within the AI Act.

Human-AI Collaboration and Explainability for 2D/3D Registration Quality Assurance

Sue Min Cho, Alexander Do, Russell H. Taylor +1

Purpose: As surgery increasingly integrates advanced imaging, algorithms, and robotics to automate complex tasks, human judgment of system correctness remains a vital safeguard for patient safety. A critical example is 2D/3D registration, where small registration misalignments can lead to surgical errors. Current visualization strategies alone are insufficient to reliably enable humans to detect these misalignments, highlighting the need for enhanced decision-support tools. Methods: We propose the first artificial intelligence (AI) model tailored to 2D/3D registration quality assessment, augmented with explainable AI (XAI) mechanisms to clarify the model's predictions. Using both objective measures (e.g., accuracy, sensitivity, precision, specificity) and subjective evaluations (e.g., workload, trust, and understanding), we systematically compare decision-making across four conditions: AI-only, Human-only, Human+AI, and Human+XAI. Results: The AI-only condition achieved the highest accuracy, whereas collaborative paradigms (Human+AI and Human+XAI) improved sensitivity, precision, and specificity compared to standalone approaches. Participants experienced significantly lower workload in collaborative conditions relative to the Human-only condition. Moreover, participants reported a greater understanding of AI predictions in the Human+XAI condition than in Human+AI, although no significant differences were observed between the two collaborative paradigms in perceived trust or workload. Conclusion: Human-AI collaboration can enhance 2D/3D registration quality assurance, with explainability mechanisms improving user understanding. Future work should refine XAI designs to optimize decision-making performance and efficiency. Extending both the algorithmic design and human-XAI collaboration elements holds promise for more robust quality assurance of 2D/3D registration.

In the era of synthetic media, deepfake manipulations pose a significant threat to information integrity. To address this challenge, we propose TrustDefender, a two-stage framework comprising (i) a lightweight convolutional neural network (CNN) that detects deepfake imagery in real-time extended reality (XR) streams, and (ii) an integrated succinct zero-knowledge proof (ZKP) protocol that validates detection results without disclosing raw user data. Our design addresses both the computational constraints of XR platforms while adhering to the stringent privacy requirements in sensitive settings. Experimental evaluations on multiple benchmark deepfake datasets demonstrate that TrustDefender achieves 95.3% detection accuracy, coupled with efficient proof generation underpinned by rigorous cryptography, ensuring seamless integration with high-performance artificial intelligence (AI) systems. By fusing advanced computer vision models with provable security mechanisms, our work establishes a foundation for reliable AI in immersive and privacy-sensitive applications.

The Endless Tuning is a design method for a reliable deployment of artificial intelligence based on a double mirroring process, which pursues both the goals of avoiding human replacement and filling the so-called responsibility gap (Matthias 2004). Originally depicted in (Fabris et al. 2024) and ensuing the relational approach urged therein, it was then actualized in a protocol, implemented in three prototypical applications regarding decision-making processes (respectively: loan granting, pneumonia diagnosis, and art style recognition) and tested with such as many domain experts. Step by step illustrating the protocol, giving insights concretely showing a different voice (Gilligan 1993) in the ethics of artificial intelligence, a philosophical account of technical choices (e.g., a reversed and hermeneutic deployment of XAI algorithms) will be provided in the present study together with the results of the experiments, focusing on user experience rather than statistical accuracy. Even thoroughly employing deep learning models, full control was perceived by the interviewees in the decision-making setting, while it appeared that a bridge can be built between accountability and liability in case of damage.

As large language models (LLMs) become increasingly integrated into real-world applications, scalable and rigorous safety evaluation is essential. This paper introduces Aymara AI, a programmatic platform for generating and administering customized, policy-grounded safety evaluations. Aymara AI transforms natural-language safety policies into adversarial prompts and scores model responses using an AI-based rater validated against human judgments. We demonstrate its capabilities through the Aymara LLM Risk and Responsibility Matrix, which evaluates 20 commercially available LLMs across 10 real-world safety domains. Results reveal wide performance disparities, with mean safety scores ranging from 86.2% to 52.4%. While models performed well in well-established safety domains such as Misinformation (mean = 95.7%), they consistently failed in more complex or underspecified domains, notably Privacy & Impersonation (mean = 24.3%). Analyses of Variance confirmed that safety scores differed significantly across both models and domains (p < .05). These findings underscore the inconsistent and context-dependent nature of LLM safety and highlight the need for scalable, customizable tools like Aymara AI to support responsible AI development and oversight.

Robust XAI techniques should ideally be simultaneously deterministic, model agnostic, and guaranteed to converge. We propose MULTIDIMENSIONAL PROBLEM AGNOSTIC EXPLAINABLE AI (MUPAX), a deterministic, model agnostic explainability technique, with guaranteed convergency. MUPAX measure theoretic formulation gives principled feature importance attribution through structured perturbation analysis that discovers inherent input patterns and eliminates spurious relationships. We evaluate MUPAX on an extensive range of data modalities and tasks: audio classification (1D), image classification (2D), volumetric medical image analysis (3D), and anatomical landmark detection, demonstrating dimension agnostic effectiveness. The rigorous convergence guarantees extend to any loss function and arbitrary dimensions, making MUPAX applicable to virtually any problem context for AI. By contrast with other XAI methods that typically decrease performance when masking, MUPAX not only preserves but actually enhances model accuracy by capturing only the most important patterns of the original data. Extensive benchmarking against the state of the XAI art demonstrates MUPAX ability to generate precise, consistent and understandable explanations, a crucial step towards explainable and trustworthy AI systems. The source code will be released upon publication.

As AI systems become increasingly capable and ubiquitous, ensuring the safety of these systems is critical. However, existing safety tools often target different aspects of model safety and cannot provide full assurance in isolation, highlighting a need for integrated and composite methodologies. This paper introduces BlueGlass, a framework designed to facilitate composite AI safety workflows by providing a unified infrastructure enabling the integration and composition of diverse safety tools that operate across model internals and outputs. Furthermore, to demonstrate the utility of this framework, we present three safety-oriented analyses on vision-language models for the task of object detection: (1) distributional evaluation, revealing performance trade-offs and potential failure modes across distributions; (2) probe-based analysis of layer dynamics highlighting shared hierarchical learning via phase transition; and (3) sparse autoencoders identifying interpretable concepts. More broadly, this work contributes foundational infrastructure and findings for building more robust and reliable AI systems.

Research into explainable artificial intelligence (XAI) for data analysis tasks suffer from a large number of contradictions and lack of concrete design recommendations stemming from gaps in understanding the tasks that require AI assistance. In this paper, we drew on multiple fields such as visual analytics, cognition, and dashboard design to propose a method for categorising and comparing XAI studies under three dimensions: what, why, and who. We identified the main problems as: inadequate descriptions of tasks, context-free studies, and insufficient testing with target users. We propose that studies should specifically report on their users' domain, AI, and data analysis expertise to illustrate the generalisability of their findings. We also propose study guidelines for designing and reporting XAI tasks to improve the XAI community's ability to parse the rapidly growing field. We hope that our contribution can help researchers and designers better identify which studies are most relevant to their work, what gaps exist in the research, and how to handle contradictory results regarding XAI design.

The rapid advancement of large language models (LLMs) has raised concerns about reliably detecting AI-generated text. Stylometric metrics work well on autoregressive (AR) outputs, but their effectiveness on diffusion-based models is unknown. We present the first systematic comparison of diffusion-generated text (LLaDA) and AR-generated text (LLaMA) using 2 000 samples. Perplexity, burstiness, lexical diversity, readability, and BLEU/ROUGE scores show that LLaDA closely mimics human text in perplexity and burstiness, yielding high false-negative rates for AR-oriented detectors. LLaMA shows much lower perplexity but reduced lexical fidelity. Relying on any single metric fails to separate diffusion outputs from human writing. We highlight the need for diffusion-aware detectors and outline directions such as hybrid models, diffusion-specific stylometric signatures, and robust watermarking.