Research Papers

The upheaval brought by the arrival of the COVID-19 pandemic has continued to bring fresh challenges over the past two years. During this COVID-19 pandemic, there has been a need for rapid identification of infected patients and specific delineation of infection areas in computed tomography (CT) images. Although deep supervised learning methods have been established quickly, the scarcity of both image-level and pixel-level labels as well as the lack of explainable transparency still hinder the applicability of AI. Can we identify infected patients and delineate the infections with extreme minimal supervision? Semi-supervised learning has demonstrated promising performance under limited labelled data and sufficient unlabelled data. Inspired by semi-supervised learning, we propose a model-agnostic calibrated pseudo-labelling strategy and apply it under a consistency regularization framework to generate explainable identification and delineation results. We demonstrate the effectiveness of our model with the combination of limited labelled data and sufficient unlabelled data or weakly-labelled data. Extensive experiments have shown that our model can efficiently utilize limited labelled data and provide explainable classification and segmentation results for decision-making in clinical routine. The code is available at https://github.com/ayanglab/XAI COVID-19.

Trustworthy Artificial Intelligence and Process Mining: Challenges and Opportunities

Andrew Pery, Majid Rafiei, Michael Simon +1

The premise of this paper is that compliance with Trustworthy AI governance best practices and regulatory frameworks is an inherently fragmented process spanning across diverse organizational units, external stakeholders, and systems of record, resulting in process uncertainties and in compliance gaps that may expose organizations to reputational and regulatory risks. Moreover, there are complexities associated with meeting the specific dimensions of Trustworthy AI best practices such as data governance, conformance testing, quality assurance of AI model behaviors, transparency, accountability, and confidentiality requirements. These processes involve multiple steps, hand-offs, re-works, and human-in-the-loop oversight. In this paper, we demonstrate that process mining can provide a useful framework for gaining fact-based visibility to AI compliance process execution, surfacing compliance bottlenecks, and providing for an automated approach to analyze, remediate and monitor uncertainty in AI regulatory compliance processes.

Artificial Intelligence (AI) systems are being deployed around the globe in critical fields such as healthcare and education. In some cases, expert practitioners in these domains are being tasked with introducing or using such systems, but have little or no insight into what data these complex systems are based on, or how they are put together. In this paper, we consider an AI system from the domain practitioner's perspective and identify key roles that are involved in system deployment. We consider the differing requirements and responsibilities of each role, and identify a tension between transparency and privacy that needs to be addressed so that domain practitioners are able to intelligently assess whether a particular AI system is appropriate for use in their domain.

Closing the AI Knowledge Gap

Ziv Epstein, Blakeley H. Payne, Judy Hanwen Shen +6

AI researchers employ not only the scientific method, but also methodology from mathematics and engineering. However, the use of the scientific method - specifically hypothesis testing - in AI is typically conducted in service of engineering objectives. Growing interest in topics such as fairness and algorithmic bias show that engineering-focused questions only comprise a subset of the important questions about AI systems. This results in the AI Knowledge Gap: the number of unique AI systems grows faster than the number of studies that characterize these systems' behavior. To close this gap, we argue that the study of AI could benefit from the greater inclusion of researchers who are well positioned to formulate and test hypotheses about the behavior of AI systems. We examine the barriers preventing social and behavioral scientists from conducting such studies. Our diagnosis suggests that accelerating the scientific study of AI systems requires new incentives for academia and industry, mediated by new tools and institutions. To address these needs, we propose a two-sided marketplace called TuringBox. On one side, AI contributors upload existing and novel algorithms to be studied scientifically by others. On the other side, AI examiners develop and post machine intelligence tasks designed to evaluate and characterize algorithmic behavior. We discuss this market's potential to democratize the scientific study of AI behavior, and thus narrow the AI Knowledge Gap.

Foundation model developers are among the world's most important companies. As these companies become increasingly consequential, how do their transparency practices evolve? The 2025 Foundation Model Transparency Index is the third edition of an annual effort to characterize and quantify the transparency of foundation model developers. The 2025 FMTI introduces new indicators related to data acquisition, usage data, and monitoring and evaluates companies like Alibaba, DeepSeek, and xAI for the first time. The 2024 FMTI reported that transparency was improving, but the 2025 FMTI finds this progress has deteriorated: the average score out of 100 fell from 58 in 2024 to 40 in 2025. Companies are most opaque about their training data and training compute as well as the post-deployment usage and impact of their flagship models. In spite of this general trend, IBM stands out as a positive outlier, scoring 95, in contrast to the lowest scorers, xAI and Midjourney, at just 14. The five members of the Frontier Model Forum we score end up in the middle of the Index: we posit that these companies avoid reputational harms from low scores but lack incentives to be transparency leaders. As policymakers around the world increasingly mandate certain types of transparency, this work reveals the current state of transparency for foundation model developers, how it may change given newly enacted policy, and where more aggressive policy interventions are necessary to address critical information deficits.

With origins in game theory, probabilistic values like Shapley values, Banzhaf values, and semi-values have emerged as a central tool in explainable AI. They are used for feature attribution, data attribution, data valuation, and more. Since exact computation can require exponential time, research has focused on efficient approximation methods using two techniques: Monte Carlo sampling and linear regression formulations. In this work, we present a new way of combining both of these techniques. Our approach is more flexible than prior algorithms, allowing for linear regression to be replaced with any function family whose probabilistic values can be computed efficiently. This allows us to harness the accuracy of tree-based models like XGBoost; with cross-fitting, the method also produces unbiased estimates. From experiments across eight datasets, we find that our methods give state-of-the-art performance for estimating probabilistic values. For Shapley values, the error of our methods can be $11\times$ lower than Permutation SHAP (the most popular Monte Carlo method), $6.5\times$ lower than Kernel SHAP (the most popular linear regression method), and $3.4\times$ lower than Leverage SHAP (the prior state-of-the-art Shapley value estimator). For more general probabilistic values, we can obtain error $390\times$ lower than the best estimator from prior work.

Examining the Expanding Role of Synthetic Data Throughout the AI Development Pipeline

Shivani Kapania, Stephanie Ballard, Alex Kessler +1

Alongside the growth of generative AI, we are witnessing a surge in the use of synthetic data across all stages of the AI development pipeline. It is now common practice for researchers and practitioners to use one large generative model (which we refer to as an auxiliary model) to generate synthetic data that is used to train or evaluate another, reconfiguring AI workflows and reshaping the very nature of data. While scholars have raised concerns over the risks of synthetic data, policy guidance and best practices for its responsible use have not kept up with these rapidly evolving industry trends, in part because we lack a clear picture of current practices and challenges. Our work aims to address this gap. Through 29 interviews with AI practitioners and responsible AI experts, we examine the expanding role of synthetic data in AI development. Our findings reveal how auxiliary models are now widely used across the AI development pipeline. Practitioners describe synthetic data as crucial for addressing data scarcity and providing a competitive edge, noting that evaluation of generative AI systems at scale would be infeasible without auxiliary models. However, they face challenges controlling the outputs of auxiliary models, generating data that accurately depict underrepresented groups, and scaling data validation practices that are based primarily on manual inspection. We detail general limitations of and ethical considerations for synthetic data and conclude with a proposal of concrete steps towards the development of best practices for its responsible use.

WalledEval is a comprehensive AI safety testing toolkit designed to evaluate large language models (LLMs). It accommodates a diverse range of models, including both open-weight and API-based ones, and features over 35 safety benchmarks covering areas such as multilingual safety, exaggerated safety, and prompt injections. The framework supports both LLM and judge benchmarking and incorporates custom mutators to test safety against various text-style mutations, such as future tense and paraphrasing. Additionally, WalledEval introduces WalledGuard, a new, small, and performant content moderation tool, and two datasets: SGXSTest and HIXSTest, which serve as benchmarks for assessing the exaggerated safety of LLMs and judges in cultural contexts. We make WalledEval publicly available at https://github.com/walledai/walledeval.

Ethical Framework for Responsible Foundational Models in Medical Imaging

Debesh Jha, Gorkem Durak, Abhijit Das +33

The emergence of foundational models represents a paradigm shift in medical imaging, offering extraordinary capabilities in disease detection, diagnosis, and treatment planning. These large-scale artificial intelligence systems, trained on extensive multimodal and multi-center datasets, demonstrate remarkable versatility across diverse medical applications. However, their integration into clinical practice presents complex ethical challenges that extend beyond technical performance metrics. This study examines the critical ethical considerations at the intersection of healthcare and artificial intelligence. Patient data privacy remains a fundamental concern, particularly given these models' requirement for extensive training data and their potential to inadvertently memorize sensitive information. Algorithmic bias poses a significant challenge in healthcare, as historical disparities in medical data collection may perpetuate or exacerbate existing healthcare inequities across demographic groups. The complexity of foundational models presents significant challenges regarding transparency and explainability in medical decision-making. We propose a comprehensive ethical framework that addresses these challenges while promoting responsible innovation. This framework emphasizes robust privacy safeguards, systematic bias detection and mitigation strategies, and mechanisms for maintaining meaningful human oversight. By establishing clear guidelines for development and deployment, we aim to harness the transformative potential of foundational models while preserving the fundamental principles of medical ethics and patient-centered care.

The black-box nature of deep learning models has raised concerns about their interpretability for successful deployment in real-world clinical applications. To address the concerns, eXplainable Artificial Intelligence (XAI) aims to provide clear and understandable explanations of the decision-making process. In the medical domain, concepts such as attributes of lesions or abnormalities serve as key evidence for deriving diagnostic results. Existing concept-based models mainly depend on concepts that appear independently and require fine-grained concept annotations such as bounding boxes. However, a medical image usually contains multiple concepts, and the fine-grained concept annotations are difficult to acquire. In this paper, we aim to interpret representations in deep neural networks by aligning the axes of the latent space with known concepts of interest. We propose a novel Concept-Attention Whitening (CAW) framework for interpretable skin lesion diagnosis. CAW is comprised of a disease diagnosis branch and a concept alignment branch. In the former branch, we train a convolutional neural network (CNN) with an inserted CAW layer to perform skin lesion diagnosis. The CAW layer decorrelates features and aligns image features to conceptual meanings via an orthogonal matrix. In the latter branch, the orthogonal matrix is calculated under the guidance of the concept attention mask. We particularly introduce a weakly-supervised concept mask generator that only leverages coarse concept labels for filtering local regions that are relevant to certain concepts, improving the optimization of the orthogonal matrix. Extensive experiments on two public skin lesion diagnosis datasets demonstrated that CAW not only enhanced interpretability but also maintained a state-of-the-art diagnostic performance.

The advent of advanced AI underscores the urgent need for comprehensive safety evaluations, necessitating collaboration across communities (i.e., AI, software engineering, and governance). However, divergent practices and terminologies across these communities, combined with the complexity of AI systems-of which models are only a part-and environmental affordances (e.g., access to tools), obstruct effective communication and comprehensive evaluation. This paper proposes a framework for AI system evaluation comprising three components: 1) harmonised terminology to facilitate communication across communities involved in AI safety evaluation; 2) a taxonomy identifying essential elements for AI system evaluation; 3) a mapping between AI lifecycle, stakeholders, and requisite evaluations for accountable AI supply chain. This framework catalyses a deeper discourse on AI system evaluation beyond model-centric approaches.

A New Perspective on Evaluation Methods for Explainable Artificial Intelligence (XAI)

Timo Speith, Markus Langer

Within the field of Requirements Engineering (RE), the increasing significance of Explainable Artificial Intelligence (XAI) in aligning AI-supported systems with user needs, societal expectations, and regulatory standards has garnered recognition. In general, explainability has emerged as an important non-functional requirement that impacts system quality. However, the supposed trade-off between explainability and performance challenges the presumed positive influence of explainability. If meeting the requirement of explainability entails a reduction in system performance, then careful consideration must be given to which of these quality aspects takes precedence and how to compromise between them. In this paper, we critically examine the alleged trade-off. We argue that it is best approached in a nuanced way that incorporates resource availability, domain characteristics, and considerations of risk. By providing a foundation for future research and best practices, this work aims to advance the field of RE for AI.