Over the past two decades, several studies have measured olfactory performance in Mild Cognitive Impairment (MCI). Deficits are observed in multiple olfactory domains, including odour detection threshold, identification, discrimination, and memory. In this study, the psychophysiological Sniffin’ Sticks smell screening test was administered to examine olfactory functioning in 145 older adults with MCI, a group with MCI and chronic comorbid conditions, and a healthy age-matched comparison group. We hypothesised that olfactory performance will deteriorate in the two MCI groups compared to the control group, even after assessing the known contributions of age and gender. The higher olfactory deterioration in the group with the MCI and the comorbidities in the first year disappeared in the second. This could mean that early consideration of the potential effect of other comorbidities that might affect olfaction should be taken and addressed, as they could easily mask the effect of cognitive decline and/or contribute to it. This study also found higher deterioration in smell identification in participants with MCI, as has been found repeatedly in similar research. Olfactory identification seems to be a more robust marker for discriminating people with MCI and without, and even discriminating between those with MCI and having other health problems.

This chapter outlines key priorities and open research directions for advancing AI-enhanced democratic deliberation, drawing on comparative insights from multiple European research initiatives. It foregrounds the need to assess the effectiveness and societal impact of AI tools beyond technical performance, with particular attention to democratic values such as inclusiveness, transparency, deliberative quality, and civic empowerment. The chapter addresses persistent challenges related to appropriateness, bias, representational fairness, and the risk of echo chambers in AI-mediated participation. It further explores hybrid human–AI configurations as socio-technical arrangements that can enrich collective reasoning while preserving human agency and institutional responsibility. Particular emphasis is placed on the evolving role of the public sector, including the strategic use of AI in policymaking and eGovernment, and on the necessity of interpretability, trustworthiness, and reliability as conditions for democratic legitimacy and regulatory alignment. The chapter concludes by reflecting on structural and methodological limitations and by identifying future research directions to support responsible, scalable, and value-aligned innovation in AI-supported democratic governance.

This chapter examines the evolving relationships between data, AI, and democracy, exploring how emerging technologies are transforming governance and civic participation. While AI holds substantial promise—enhancing public service delivery, facilitating citizen engagement, and streamlining communication between institutions and constituents—it also introduces significant risks, including algorithmic bias, misinformation, surveillance, and corporate dominance. These dual dynamics frame the chapter’s analysis, which is anchored in both theoretical foundations and empirical observations. In light of this, the chapter begins by unpacking the promise and risk associated with AI in democratic contexts, situating current debates within the broader landscape of civic technologies and e-participation. It then examines the state of the art through two key sources: the 2025 Digital Participation Tools Ratings by People Powered and the IPU's 2025 report on AI use in parliaments. Finally, the chapter situates four EU-funded initiatives—ORBIS, KT4D, AI4GOV, and ITHACA—within this landscape, highlighting their contributions to advancing democratic innovation through technological, ethical, and participatory approaches.