Keynote Talks

The New Scarcity: Human Judgment in a World of AI-Generated Content

Edgar Weippl
University of Vienna, Austria

Abstract

Large language models make it cheap to produce polished text, code, applications, reviews, objections, and complaints. The bottleneck, therefore, moves from production to human judgment. Journals, grant agencies, universities, companies, and public administrations may face a denial-of-service problem at the human and organizational layer: too many plausible artifacts requiring careful assessment. This keynote frames the issue as layer-8 denial-of-service and asks how institutions can protect human attention without sacrificing openness, fairness, and privacy. I discuss proof of human time, rate limits, accountable pseudonyms, and proof of useful work as possible building blocks for resilient institutions.

Biography Edgar Weippl is a Full Professor for Security and Privacy at the University of Vienna and Vice Dean of the Faculty of Computer Science. He is also the Research Director of SBA Research and a visiting professor at the National Institute of Informatics in Tokyo. His research areas include software security. Edgar is a senior associate editor of IEEE Transactions on Information Forensics and Security, associate editor for ACM Transactions on Privacy and Security, and IFIP TC11 liaison to Computers & Security. He is Austria’s representative at IFIP TC 11 for Security and Privacy in Information Processing Systems.

Phishing Attacks and the Paradox of Privacy in Data-Driven Digital Ecosystems

Anne Kayem
University of Exeter, UK

Abstract

Data constitutes a central driver of service-oriented digital ecosystems. However, its use raises a range of ethical questions, particularly regarding the modalities of data collection, the conditions under which data are stored and managed, and the mechanisms through which data privacy is safeguarded.
The growing complexity of ensuring adequate privacy protection has led to the emergence of Privacy-as-a-Service (PaaS) platforms, through which organisations outsource privacy risk assessment and regulatory compliance activities to specialised third-party providers. This approach reduces the necessity for dedicated in-house expertise and resources, and can improve the cost–benefit ratio associated with preventing data breaches and avoiding regulatory sanctions.
Nonetheless, in view of the increasing prevalence and sophistication of data-driven cyber security incidents, it is evident that additional scrutiny and more robust mechanisms are required to provide strong and verifiable guarantees of data protection.
In this presentation, I examine three interrelated dimensions of data protection and privacy within data-driven digital ecosystems. 
The first concerns the ways in which legal or technical data protection guarantees may come into tension with prevailing conceptions of digital privacy. Safeguarding digital privacy is particularly challenging for two principal reasons: (1) end-users frequently lack a comprehensive understanding of which specific behaviours and interactions precipitate privacy loss, as well as the mechanisms available to mitigate such loss; and (2) service providers depend on extensive data collection and processing to ensure the effective and efficient functioning of their applications, thereby obscuring the boundary between data that is strictly necessary and data that is merely desirable. In addition, practices of hyper-personalisation, increasingly facilitated by generative artificial intelligence systems, introduce further layers of complexity for the assurance of privacy and, by implication, for the robustness of data protection mechanisms.
The second aspect concerns the increasing availability and widespread adoption of ready-made tools that rely on generative artificial intelligence to lower the entry barrier to cybercrime, particularly phishing. Data from 2025 indicates that phishing-as-a-service (PhaaS) platforms offer complete phishing kits for prices starting at approximately USD 40. Phishing attacks fundamentally exploit a data-as-a-service model, whereby personal information is exfiltrated from users—constituting a violation of informational privacy—and subsequently utilised to execute a range of cyberattacks (e.g. unauthorised data processing, manipulation or fabrication of consent, and related forms of data misuse) that effectively circumvent data protection safeguards.
The third aspect pertains to the normative and conceptual foundations of employing Privacy-as-a-Service platforms. In particular, it concerns the consequences of non-compliance with data protection and privacy legislation in scenarios involving the compromise of sensitive personal data, including large-scale data breaches. In this context, we examine, as an illustrative example, deepfake-based phishing attacks and analyse their implications for data protection, regulatory accountability, and the robustness of existing privacy-preserving frameworks.

Biography Anne Kayem is an Associate Professor of Cyber-Security and leads the Privacy AnaLytics (PAL) research group at the University of Exeter. Her research is situated within the broader domain of database security, with a particular focus on digital privacy. In this context, she concentrates on the design and analysis of data transformation algorithms that facilitate privacy-preserving machine learning and data analytics. Her work examines the theoretical and practical underpinnings of mechanisms required to conceptualise, formalise, and implement privacy-preserving solutions that account for both computational constraints (e.g., efficiency, scalability, and resource limitations) and human-centric factors (e.g., usability, interpretability, and compliance with regulatory and ethical frameworks). She has authored and edited several scholarly books and book chapters, and has published extensively in peer-reviewed journals and conference proceedings in the fields of cyber-security and privacy. Her contributions encompass areas such as Access Control, Information Security, secure micro-grids, and, more recently, Digital Privacy, where she investigates methods for safeguarding sensitive information in complex, data-intensive environments. She is actively engaged in editorial and peer-review activities for leading venues, including journals such as Elsevier Computers & Security and Springer Information Security, as well as a range of international conferences in cyber and data security. She is a Senior Member of both the Association for Computing Machinery (ACM) and the Institute of Electrical and Electronics Engineers (IEEE). In addition, she participates actively in professional organisations that promote gender equity and support women in academia, including ACM-W, IEEE Women in Engineering (WIE), and UK Women in Higher Education (WHEN).

Community Mining and Community Search: Beyond Social Network Analysis

Osmar Zaïane
University of Alberta, Canada

Abstract

Community mining has long been associated with the analysis of social networks, where the goal is to discover groups of closely connected individuals. Today, however, the scope of community mining extends far beyond social interactions. Modern applications involve complex networks whose nodes carry rich semantic information and whose relationships are often uncertain or probabilistic, reflecting the noisy and incomplete nature of real-world data.
This talk will provide an accessible introduction to community mining and community search, beginning with their foundations in graph analysis before exploring recent advances for attributed and probabilistic graphs. We will discuss how uncertainty and attributes fundamentally change the notion of a community and motivate the development of new algorithms for both community detection and community search. As an illustration, we will present USIWO, a local community search algorithm designed for uncertain graphs that efficiently identifies high-quality communities while explicitly accounting for probabilistic relationships.  
The second half of the talk will highlight a broad spectrum of emerging applications, with a particular emphasis on artificial intelligence. We will see how community mining can organize knowledge graphs, improve retrieval-augmented generation (RAG), support long-term memory in conversational AI, detect hallucinations through confidence-aware knowledge structures, and facilitate collaboration in multi-agent AI systems. Additional examples from bioinformatics, cybersecurity, recommender systems, and scientific knowledge discovery will illustrate how community mining has evolved into a powerful paradigm for extracting structure from complex, uncertain, and heterogeneous data.
By looking beyond its origins in social network analysis, this talk will demonstrate how community mining is becoming a key enabling technology for the next generation of intelligent, trustworthy, and explainable AI systems.

Biography Osmar Zaïane is a Professor in Computing Science at the University of Alberta, Canada, Fellow of the Alberta Machine Intelligence Institute (Amii), and Canada CIFAR AI Chair. Dr. Zaiane obtained his Ph.D. from Simon Fraser University, Canada, in 1999. He has published more than 400 papers in refereed international conferences and journals. He is an Associate Editor of many International Journals on data mining and data analytics and served as program chair and general chair for scores of international conferences in the field of knowledge discovery and data mining. Dr. Zaiane received numerous awards including the 2010 ACM SIGKDD Service Award from the ACM Special Interest Group on Data Mining, which runs the world’s premier data science, big data, and data mining association and conference.  Dr. Zaiane focuses on pattern discovery and information extraction from large databases, also known as data mining. His work involves data mining from disparate heterogeneous data sources, such as on the Internet, as well as the analysis of complex information networks, also known as social network analysis. Specific research projects include the development of tools for data analytics such as Meerkat, a tool for analyzing changes over time in a network of entities. Other of Dr. Zaiane’s research projects relate to data mining in health informatics and the development of tools for document categorization and decision support systems. He focuses on building applications that can improve decision-making in fields from business to medicine, allowing decisions to be based on data and data analysis. Through the application of machine learning and methods of knowledge discovery, he devises ways to personalize applications, automate processes and improve upon current data science practices.