AI Ethics: An Empirical Study on the Views of Practitioners and Lawmakers

Abstract

Artificial intelligence (AI) solutions and technologies are being increasingly adopted in smart systems contexts; however, such technologies are concerned with ethical uncertainties. Various guidelines, principles, and regulatory frameworks are designed to ensure that AI technologies adhere to ethical well-being. However, the implications of AI ethics principles and guidelines are still being debated. To further explore the significance of AI ethics principles and relevant challenges, we conducted a survey of 99 randomly selected representative AI practitioners and lawmakers (e.g., AI engineers and lawyers) from 20 countries across five continents. To the best of our knowledge, this is the first empirical study that unveils the perceptions of two different types of population (AI practitioners and lawmakers) and the study findings confirm that transparency, accountability, and privacy are the most critical AI ethics principles. On the other hand, lack of ethical knowledge, no legal frameworks, and lacking monitoring bodies are found to be the most common AI ethics challenges. The impact analysis of the challenges across principles reveals that conflict in practice is a highly severe challenge. Moreover, the perceptions of practitioners and lawmakers are statistically correlated with significant differences for particular principles (e.g. fairness and freedom) and challenges (e.g. lacking monitoring bodies and machine distortion). Our findings stimulate further research, particularly empowering existing capability maturity models to support ethics-aware AI systems' development and quality assessment.

Trayectoria y modelo de gobernanza de las políticas de inteligencia artificial (IA) de los países de América del Norte

“…Ante los efectos transformadores que tendrá la IA, se ha vuelto vital para las sociedades el delimitar la estrategia de parámetros relevantes para construir entornos de IA seguros, en este contexto que hace poco más de un lustro surgen conceptos como la ética y la gobernanza de la IA. De acuerdo con Khan et al (2023Khan et al ( , p. 2972) la ética de la IA se relaciona con las preocupaciones relacionadas con los efectos sociales y el potencial transformador de las tecnologías que utilicen algoritmos de IA, que pueden relacionarse con temas como la autonomía humana, la protección de derechos como la privacidad, la libertad en una sociedad democrática y la calidad de vida. De esta forma, se vuelve necesaria la reflexión ética en las tecnologías de IA para fomentar valores éticos alineadas al desarrollo social, la protección de los derechos humanos y las garantías de los individuos, en el marco del desarrollo de esta tecnología (Dignum, 2017, p.5).…”

“…Sobre este punto Khan et al (2023) a través de la revisión de un total de 27 estudios identificaron un total de los 22 principios éticos que pueden apoyar en la construcción de la gobernanza de la IA. Del mismo modo, Dafoe (2018) indica que principios como la rendición de cuentas (accountability), la justicia (fairness) o la transparencia (transparency).…”

A trustworthy AI reality-check: the lack of transparency of artificial intelligence products in healthcare

“…So far, the limited documentation of medical AI software has been justified by a lack of understanding on ethics among developers ( 51 – 53 ) or as a threat to intellectual property (IP) ( 30 , 54 ). Another reason for these identified transparency gaps may be that transparency as a principle for trustworthy AI is only vaguely defined ( 19 , 55 ).…”

Trust, artificial intelligence and software practitioners: an interdisciplinary agenda

“…Trust and trustworthiness are central concepts in contemporary discussions about Artificial Intelligence (AI)-based software and the relationships between people and AI, with an expansive body of literature and debate relating to human trust in AI and automated systems and technologies (Shahrdar et al 2019). There is also significant existing research regarding how concepts of trust and trustworthiness are used in industry and policy domains (Khan et al 2023). However, as existing critiques reveal (Pink 2022(Pink , 2023; Reinhardt 2022, Shneiderman 2020) both concepts have been defined and used differently across software engineering and computer science disciplines (Shneiderman 2020;Lu et al 2022a, b;Ahmad et al 2023), in agendas of industry and government, in the humanities and in the qualitative social sciences (Pink 2022;Reinhardt 2022).…”

“…Seaver 2022, Lanzeni andArdevol 2017), and this point is equally applicable to AI. However, little is known about how software engineers who design and develop AI-based software applications conceptualise trust in AI as they go about their everyday work practices or about the implications of this for sociotechnical relationships (Khan et al 2023;Ahmad et al 2023).…”