Nuno Marques

António Branco, Luís Gomes, Rodrigo Santos et al.

The rapid expansion of AI-based remote services has intensified debates about the long-term implications of growing structural concentration in infrastructure and expertise. As AI capabilities become increasingly intertwined with geopolitical interests, the availability and reliability of foundational AI services can no longer be taken for granted. This issue is particularly pressing for AI-enabled public services for citizens, as governments and public agencies are progressively adopting 24/7 AI-driven support systems typically operated through commercial offerings from a small oligopoly of global technology providers. This paper challenges the prevailing assumption that general-purpose architectures, offered by these providers, are the optimal choice for all application contexts. Through practical experimentation, we demonstrate that viable and cost-effective alternatives exist. Alternatives that align with principles of digital and cultural sovereignty. Our findings provide an empirical illustration that sovereign AI-based public services are both technically feasible and economically sustainable, capable of operating effectively on premises with modest computational and financial resources while maintaining cultural and digital autonomy. The technical insights and deployment lessons reported here are intended to inform the adoption of similar sovereign AI public services by national agencies and governments worldwide.

António Branco, João Silva, Nuno Marques et al.

Although the cultural (mis)alignment of Large Language Models (LLMs) has attracted increasing attention -- often framed in terms of cultural bias -- until recently there has been limited work on the design and development of datasets for cultural assessment. Here, we review existing approaches to such datasets and identify their main limitations. To address these issues, we propose design guidelines for annotators and report on the construction of a dataset built according to these principles. We further present a series of contrastive experiments conducted with this dataset. The results demonstrate that our design yields test sets with greater discriminative power, effectively distinguishing between models specialized for a given culture and those that are not, ceteris paribus.

Nuno Marques, André Zúquete, João Paulo Barraca

In a scenario where hotspot wireless networks are increasingly being used, and given the amount of sensitive information exchanged on Internet interactions, there is the need to implement security mechanisms that guarantee data confidentiality and integrity in such networks, as well as the authenticity of the hotspot providers. However, many hotspots today use Captive Portals, which rely on authentication through Web pages (thus, an application-level authentication approach) instead of a link-layer approach. The consequence of this is that there is no security in the wireless link to the hotspot (it has to be provided at upper protocol layers), and is cumbersome to manage wireless access profiles (we need special applications or browsers' add-ons to do that). This work exposes the weaknesses of the Captive Portals' paradigm, which does not follow a unique nor standard approach, and describes a solution that intends to suppress them, based on the 802.1X architecture. This solution uses a new EAP-compliant protocol that is able to integrate an HTTP-based registration or authentication with a Captive Portal within the 802.1X authentication framework.