2026· Annales Mathematicae et Informaticae· 0 citations· 20 references
TL;DR
A baseline review of the selected papers volume of the inaugural International Conference on Formal Methods and Foundations of Artificial Intelligence (FMF-AI 2025), published as Annales Mathematicae et Informaticae, introduces simple baseline metrics that can be recomputed in future FMF-AI editions to observe structural changes in the research landscape.
Abstract
. Artificial Intelligence (AI) is advancing rapidly, yet many successful models remain opaque and provide limited assurance about reliability, safety, and failure modes. This motivates renewed interest in formal methods and foundational perspectives that can support trustworthy AI beyond empirical testing. This paper presents a baseline review of the selected papers volume of the inaugural International Conference on Formal Methods and Foundations of Artificial Intelligence (FMF-AI 2025), published as Annales Mathematicae et Informaticae , Vol. 61 (2025). The goal is twofold: (i) to map and summarize the first FMF-AI “snapshot” as a starting point for the Hungarian research ecosystem, and (ii) to define a reproducible baseline that can serve as a reference for measuring topical and methodological shifts in subsequent FMF-AI editions. The review clusters the twenty selected papers into five thematic groups and records their relative prevalence. In addition, it introduces simple baseline metrics that can be recomputed in future FMF-AI editions to observe structural changes in the research landscape. The main pattern is a clear imbalance between verification-oriented contributions and papers that primarily use AI methods in application or optimization contexts.
This paper demonstrates an AI-assisted search process to aid in literature surveys within a fast-moving research area, and assesses the usefulness and validity of these results.
Jonathan P. Bowen, Sin-Hao Chen· Applied Sciences· 0 citations
XAI provides a powerful framework for responsible AI development, challenges such as the performance-interpretability trade-off, lack of standardized evaluation metrics, and potential for human misinterpretation remain areas of active research.
P. Pradhan, Amol Rajmane, C. patil· Journal of image processing...· 0 citations
Over the past decade, the exponential integration of artificial intelligence (AI) systems across various sectors has been propelled by significant advances in machine learning algorithms, data availability, and computational power. This progress has produced highly effective AI systems, but also underscores the critical need for effective auditing to critically evaluate these technologies. In this paper, we conduct a systematic review of the literature on methodologies, frameworks, and techniques for auditing AI systems, focusing on legal and ethical considerations and compliance with regulations. By reviewing key academic databases, including Google Scholar, IEEE, ACM, and Springer, we establish the scope of our survey and derive topics from our research questions. Our findings reveal gaps in current auditing practices and highlight the importance of incorporating AI value chain stages and AI maturity levels into auditing frameworks. This approach enables us to distinguish and recommend existing frameworks and methodologies that are most suitable for the specific contexts of different organisations, thus enhancing the effectiveness of AI system evaluations.
Usman Shahbaz, Amin Beheshti, B. Abedin et al.· ACM Computing Surveys· 0 citations
A framework for sustainable, human-centered integration of AI is proposed in which AI is restricted to technical verification and efficiency, while judgments on scientific merit, ethics, and paradigm-shifting research are reserved for appropriately valued human experts.
This work presents a comprehensive literature review of the current state of the subfield of XAI that consist of causality-motivated post-hoc XAI methods, and a causal framework for categorising XAI is introduced, and three types of post-hoc XAI methods are identified: observational methods, internally causal methods and externally causal methods.
Anna Rodum Bjøru, Helge Langseth, Inga Strümke et al.· Machine-mediated learning· 1 citation
Drawing upon recent academic publications and reports, this paper systematically examines current developments in artificial intelligence (AI) through a five-stage framework encompassing technology, applications, expectations and reality, risks and safety, and control and governance. Based on this analysis, the paper identifies key challenges and proposes future directions for AI research and development.
Following the emergence of ChatGPT, large language models (LLMs) have rapidly proliferated. Their core component, the transformer, has expanded beyond natural language processing into diverse domains such as computer vision (Vision Transformer, ViT) and multivariate time-series analysis (Time Series Transformer, TST). Furthermore, LLMs are evolving into agent-based systems and are being applied to the automation of scientific research, as exemplified by Google’s Co-Scientist, AlphaEvolve, and AlphaGenome. These developments have significantly heightened expectations regarding the transformative potential of AI across society.
However, a gap remains between these expectations and the current state of technology, as structural issues such as data bias and hallucination continue to pose substantial risks. In particular, hallucination is interpreted as a phenomenon arising from the model’s tendency to maximize expected evaluation outcomes, and it is identified as a critical challenge for ensuring AI safety.
Accordingly, the safe deployment of AI requires effective monitoring of model behavior and improved interpretability of chain-of-thought (CoT) reasoning processes, red-teaming activities at both macro- and micro-levels, and the establishment of international governance frameworks, including those in the healthcare domain such as guidelines from the World Health Organization (WHO).
In conclusion, while AI is driving profound changes not only in science and technology but also across society as a whole, addressing technical challenges— such as mitigating hallucination, preventing catastrophic forgetting in continual learning, and improving data efficiency—must be accompanied by the development of control and governance systems aligned with human values. In particular, international governance initiatives are needed to reduce disparities between countries and address polarization at the global level.
Sang-Hoon Oh· Liberal Arts Innovation Cent...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
