Assessing the ethical and social concerns of artificial intelligence in neuroinformatics research: an empirical test of the European Union Assessment List for Trustworthy AI (ALTAI)

Abstract

Ethical and social concerns are a key obstacle to the adoption of artificial intelligence (AI) in the life sciences and beyond. The discussion of these issues has intensified in recent years and led to a number of approaches, tools and initiatives. Key amongst them is the idea of ex-ante impact assessments that aim to identify issues at the early stages of development. One prominent example of such ex-ante impact assessment is the European Union's (EU) Assessment list for Trustworthy AI (ALTAI). This article uses the findings of a large-scale application of the ALTAI to a large neuro-informatics project as an exemplar to demonstrate the effectiveness and limitations of the ALTAI in practice. The article shows that ex-ante impact assessment has the potential to help identify and address ethical and social issues. However, they need to be understood as part of a broader socio-technical ecosystem of AI. For ALTAI and related approaches to be useful in bio-medical research, they should be interpreted from a systems theory perspective which allows for their integration into the rich set of tools, legislation and approaches. The paper argues that ex-ante impact assessments have the best chance of being successful if seen applied in conjunction with other approaches in the context of the overall AI ecosystem.

ALTAI Tool for Assessing AI-Based Technologies: Lessons Learned and Recommendations from SHAPES Pilots

“…SHAPES partners have acknowledged that the ALTAI tool is a reliable and easyto-use self-assessment tool for Trustworthy AI-based technology in pilot studies such as SHAPES pilots, which are of interest in the market. Similar analyses [11,12] have reported the importance of the ALTAI tool to provide reliable recommendations towards the improvement of trustworthiness in AI solutions. Specifically, although authors in [11] have slightly altered the ALTAI tool questions in their study, their results' analysis showed an awareness of some of the broader key areas of trustworthy factors as indicated by the ALTAI tool, such as accuracy of the performance of the AI solution.…”

“…Similar analyses [11,12] have reported the importance of the ALTAI tool to provide reliable recommendations towards the improvement of trustworthiness in AI solutions. Specifically, although authors in [11] have slightly altered the ALTAI tool questions in their study, their results' analysis showed an awareness of some of the broader key areas of trustworthy factors as indicated by the ALTAI tool, such as accuracy of the performance of the AI solution. Moreover, the study revealed that the participating researchers were unsure whether their work was covered by the various definitions and how applicable the recommendations could be to their solutions.…”

“…Additionally, the public guidelines have no coherent and consolidated approaches for technology adoption by AI developers [9] and healthcare professionals. Recent studies have highlighted that healthcare professionals' (physicians and nurses) perceptions are a critical issue in adopting AI technologies (such as AAL) in healthcare delivery; the negative perception of the trustworthiness of AI is tantamount to a rejection of AI [10][11][12][13]. bordering on respondent demographics such as age and other items centered on perceptions of AI's impact on the future role of healthcare professionals and their professional preparedness.…”

Leveraging Quantum Computing for Inclusive and Responsible Ai Development: A Conceptual and Review Framework

“…The societal impact of artificial intelligence (AI) has been a topic of increasing concern and scrutiny in recent years. Numerous studies have highlighted the ethical, legal, and socio-economic implications of AI deployment across various domains (Stahl, and Leach, 2023;Taeihagh, 2021). Ethically, AI systems raise concerns regarding bias, fairness, and accountability.…”

A Literature Review of Human–AI Synergy in Decision Making: From the Perspective of Affordance Actualization Theory

“…Moreover, to help users to improve their understanding of the inherent risks of AI systems and to mitigate these risks, organizations and scholars have created several tools. For example, the European Commission created the Assessment List for Trustworthy AI (ALTAI) [22,23], which indicates self-assessment checklist questions and a set of steps that organizations are encouraged to complete. Each question presents multiple choices to verify the implementation of the guidelines.…”