In today’s digital age, the integration of artificial intelligence (AI) into civic engagement platforms offers unprecedented opportunities to democratize decision-making processes. However, this transformational potential can only be realized through the meaningful inclusion of diverse voices, particularly those from marginalized or vulnerable communities. Recognizing this imperative, the ITHACA project embarks on a journey to bridge the gap between AI innovation and inclusive civic participation. This article delves into the outcomes of various activities conducted within Work Package 2 (WP2) of the project, shedding light on the intersection of algorithmic risk assessment and human-centered design in the selection process of AI Citizen Juries.

AI Citizen Juries Selection

At the heart of the ITHACA project lies the selection process for AI Citizen Juries, a cornerstone for inclusive decision-making. Recognizing the historical underrepresentation of vulnerable groups in civic processes, the project sets out to create a platform where diverse perspectives are not only welcomed but celebrated. The selection methodology adopts a multi-faceted approach, drawing insights from literature reviews, consultations with experts, and analysis of local data. By identifying vulnerable groups and variables constituting vulnerability, the project aims to create a jury pool that mirrors the rich tapestry of society.

Multifaceted Approach

Ensuring inclusivity within the AI Citizen Juries of the ITHACA project is not just a goal but a meticulously planned process. Drawing from a multifaceted approach, the project navigates the complexities of representation and vulnerability to create a diverse and inclusive platform for civic engagement.

The methodology begins with an extensive review of literature and governmental reports, providing insights into the various dimensions of vulnerability. Project partners, well-versed in inclusive digitalization, offer valuable perspectives, enriching the discourse on inclusion. Statistical data and discussions with local stakeholders further inform the selection process, ensuring contextual relevance and sensitivity.

Two comprehensive lists are formulated to guide the identification of vulnerable groups and variables constituting vulnerability. From elderly individuals to refugees, from people with disabilities to marginalized communities, every segment of society is considered. Variables such as income, education, and access to resources are meticulously examined to capture the nuanced facets of vulnerability.

Notably, differences in social contexts between pilot cities result in tailored refinements to the selection criteria. While the city of Brasov emphasizes factors like poor living conditions and limited access to cultural resources, the town of Martin prioritizes considerations such as physical disabilities and rural isolation. These adjustments underscore the project’s commitment to contextual relevance and inclusivity.

The methodology’s strength lies in its adaptability and responsiveness to diverse social landscapes. By recognizing and addressing the unique needs of each community, the project ensures that AI Citizen Juries truly reflect the rich tapestry of society. Through this nuanced approach, the project paves the way for meaningful civic engagement and ethical AI governance.

In conclusion, the methodology approach adopted by the ITHACA project exemplifies a holistic and inclusive approach to AI governance. By prioritizing representation and vulnerability, the project sets a precedent for inclusive decision-making processes in the realm of AI. As the project progresses, its methodology serves as a beacon of hope, guiding the way towards a future where every voice is heard and valued.

Diverse tapestry of participants

The AI Citizen Juries in Martin and Brasov offer a fascinating glimpse into the demographics and selection process, shedding light on the diverse tapestry of participants and the intricacies of vulnerability assessment.

City of Martin

In Martin, 30 individuals engaged with the AI Citizen Jury, with 25 opting for online registration and 5 utilizing printed forms. Gender distribution skewed towards females, with 20 females compared to 10 males. The age spectrum varied, with a majority falling within the 31-60 age bracket. Educational backgrounds also varied, with a significant portion having completed high school. Pensioners formed a notable portion of applicants, alongside individuals in full-time jobs. Most applicants resided in their own houses or flats, reflecting a diverse household landscape.

Brasov City

Brasov witnessed 26 online registrations, showcasing a relatively balanced gender distribution. The applicant pool skewed younger compared to Martin, with a higher proportion under 31 years old. A majority held university degrees, indicative of a higher educational attainment level. Occupationally, students, full-time workers, and pensioners constituted the majority. Housing varied, with single houses being most common.

Selection Process

Registrants underwent a vulnerability assessment, assigning scores based on various factors. This scoring system facilitated the selection of 25 citizens per city, prioritizing those with higher vulnerability scores. In Brasov, only one applicant was excluded due to a low vulnerability score, resulting in a diverse selection across genders and age groups. Conversely, in Martin, five applicants with the lowest vulnerability scores were placed on a waiting list. The selected participants exhibited a varied vulnerability score range, highlighting differences between the two cities.

The results underscore notable differences between Martin and Brasov, particularly in gender distribution, age representation, and vulnerability scores. While Brasov leaned towards a younger demographic with a balanced gender ratio, Martin showcased a higher proportion of older females. Additionally, the variance in vulnerability scores was more pronounced in Brasov, indicating a potentially higher degree of vulnerability among applicants.

In conclusion, the demographic insights and selection process outcomes provide valuable insights into the nuances of AI Citizen Juries. These findings serve as a foundation for informed decision-making and the continued pursuit of inclusive civic engagement in AI governance.


The ITHACA project’s endeavours in AI Citizen Juries selection exemplify a paradigm shift towards inclusive civic engagement. By cantering the voices of marginalized communities, the project lays the groundwork for a more equitable and participatory society. As the project continues to unfold, its insights and methodologies offer valuable lessons for fostering inclusive decision-making processes in an increasingly digital world. Through collaboration and empathy, we can pave the way towards a future where every voice is heard and valued.

Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or the Europe Research Executive Agency. Neither the European Union nor the granting authority can be held responsible for them.