AI and automated processes have been embedded in our every-day life and often we fail to realise the social impact they can have, positive or negative. Many times, these tools can lead to discrimination towards certain citizen groups and create fractions. On the other hand, non-AI tools and processes, also pose challenges especially to underrepresented groups, leading again to discrimination.
But how do traditional biases (non-AI) currently impact the rights and values of underrepresented groups? In the public sector, there is a growing need to correct the errors and provide the same quality of services to all citizens regardless of their personal characteristics. In this vein, AI4Gov conducted a study with a series of focus groups in the three AI4Gov pilot countries: Greece, Spain, and Slovenia. The aim was to gather in-depth insights on experiences of vulnerable groups of citizens when interacting with public services, either in person, or online. The study target groups were Students, Young unemployed adults, People with disabilities, Elderly individuals and Gender-based Violence Victims. The overarching goal of the study was to incorporate the perspectives and experiences of marginalised groups into the Holistic Regulatory Framework (find it in D2.2).
We collaborated with six organisations:
- IASIS at Work (Greece) is a day centre supporting workers and the unemployed, reaching out to young adults, permanent employees, and the long-time unemployed.
- EDRA Social Cooperative Activities for Vulnerable Groups (Greece) is a civil, non-profit cooperative organisation dedicated to promoting mental health services and safeguarding the rights of socially vulnerable groups.
- Aristotle University of Thessaloniki, Department of Political Sciences (Greece), bringing the aspect of young adults.
- REUSE Centre (CPU) – (Slovenia) is an epicentre of sustainability and social innovation, creating employment opportunities for marginalised populations.
- ALMA Association (Spain) is dedicated to fighting gender violence and achieving real equality between men and women.
- Autism Federation of Extremadura (FAE) – (Spain) is a federation that protects people with autism spectrum disorder, supporting also their families.
The study revealed significant challenges faced by vulnerable citizens in accessing public services, highlighting issues of discrimination and inequality. One major problem identified is the communication barriers, including unavailability and unhelpfulness of staff, as well as delays caused by bureaucratic inefficiencies. Additionally, a lack of technological proficiency among employees, particularly older ones, exacerbates these issues, highlighting the need for comprehensive training programs to enhance digital literacy and improve service delivery.
Vulnerable groups, such as victims of gender-based violence (GBV) and non-native speakers, face specific challenges. Public service employees often lack the empathy and professional skills necessary to support GBV victims adequately. For non-native speakers and non-EU citizens, there are significant barriers due to poor translation services, complex eligibility criteria, and the need to navigate services in less accessible languages. Autistic individuals and their families also encounter difficulties, particularly in accessing emergency healthcare services, highlighting a critical need for specialized training and protocols.
Inconsistent procedures and miscommunication further complicate public service delivery, leading to repeated visits and widespread frustration among citizens. This inconsistency, combined with a general lack of technological competence among public servants, particularly affects elderly and technologically inexperienced individuals. As a result, there is a pressing need for comprehensive training programs to improve service efficiency and consistency.
The lack of trust in public services is another critical issue, often stemming from unhelpful attitudes and insufficient training among staff. While some participants reported positive experiences, these were exceptions rather than the norm. This gap in support, especially for GBV victims and families with autistic members, erodes trust and highlights the need for systemic reforms.
Leveraging the insights from the case, we extracted some recommendations to guide the public services through the most problematic areas that were identified during the focus groups. The aim was to indicate the most common issues encountered by the citizens that require extra attention when building a public service tool, in order to minimise potential discriminatory incidents.
Training and Professional Development
- Implement specialized training for handling GBV Cases
- Establish specialised training for healthcare providers to support autistic individuals
- Improve technological literacy
Service delivery and Process Improvement
- Revise Face-to-Face Services to establish clear guidelines and accountability measures
- Standardise Procedures across public services
- Implement protocols for priority care for autistic individuals and their caregivers
Accessibility and Inclusivity
- Improve Translation Services and simplify the navigation of public websites
- Ensure that digital services are user-friendly and accessible to all citizens, including the elderly and technologically inexperienced
- Ensure that public servants are adequately trained and prepared to address the specific needs of vulnerable populations
Cultural Change and Public Trust
- Foster empathy and professionalism within public services
- Implement feedback mechanisms to continuously improve service delivery based on citizen experiences.
Technological Enhancement and Digital Services
- Enhance the technological literacy of public servants through regular training and support.
The above-mentioned recommendations can help the public services identify the difficulties vulnerable people face when accessing them. You can find the analytical methodology and results of the study in our deliverable report 2.2 (AI4Gov Holistic Regulatory Framework V2).
Disclaimer
The sample was not enough to be representative, and for this reason the results cannot be generalised for the countries. The purpose was to extract some initial observations on the public sector areas that are perceived as problematic. AI4Gov suggest a methodology on how to conduct similar activities and gather information from vulnerable populations.
