Trustworthy artificial intelligence

Abstract

Artificial intelligence (AI) brings forth many opportunities to contribute to the wellbeing of individuals and the advancement of economies and societies, but also a variety of novel ethical, legal, social, and technological challenges. Trustworthy AI (TAI) bases on the idea that trust builds the foundation of societies, economies, and sustainable development, and that individuals, organizations, and societies will therefore only ever be able to realize the full potential of AI, if trust can be established in its development, deployment, and use. With this article we aim to introduce the concept of TAI and its five foundational principles (1) beneficence, (2) non-maleficence, (3) autonomy, (4) justice, and (5) explicability. We further draw on these five principles to develop a data-driven research framework for TAI and demonstrate its utility by delineating fruitful avenues for future research, particularly with regard to the distributed ledger technology-based realization of TAI.

Drivers and Inhibitors for Organizations’ Intention to Adopt Artificial Intelligence as a Service

“…With this study, we tap into a new field of behavioral aspects of AIaaS research, leaving plenty of opportunities for further ongoing research (e.g., conceptualizing our findings differentiated by industry or region, or trust in AIaaS [42]). Given that AIaaS is a broad field, future research could study the adoption of specific technologies in the context of AIaaS technologies, such as explainable ML, transfer learning, or privacy-preserving learning, or the convergence of AI-aaS with further emerging technologies such as decentralized marketplaces empowered by distributed ledger technology and trusted hardware [43].…”

Artificial Intelligence as a Service

“…Mell and Grance 2011). Today, most developed, deployed, used AI-based systems are based on machine learning or deep learning methods (Pandl et al 2020;Thiebes et al 2020). As such, machine-learning-based techniques are also crucial technologies for the most popular AI software services.…”

“…Because the development and training of an AI model are expensive and time-consuming, AI models became a form of intellectual property and, therefore, increasingly represent an essential factor in achieving competitive advantages (Haenlein and Kaplan 2019). Efforts to protect competitive advantages can thus lead to situations in which promising AI models are not shared with others (Thiebes et al 2020). To counteract this issue, a type of AI software service emerged that removes users' burden of setting up and training, and offers pre-trained models, referring to AI models already trained by the AIaaS provider (or other parties) and then made available to users.…”

“…These advantages are among the most discussed benefits in prior research and practice because successfully optimizing each machine learning step requires overcoming significant complexity that is difficult without in-depth knowledge and experience (Yao et al 2017). Offering inference as a service and pre-trained models is thus an efficient means to make promising AI models more widely available to be highly beneficial to society (Thiebes et al 2020).…”

“…

is undoubtedly one of the most actively debated technologies, providing auspicious opportunities to contribute to individuals' well-being, the success and innovativeness of organizations, and societies' prosperity and advancement (Thiebes et al 2020). The McKinsey Global Institute predicts that the utilization of AI could yield an additional worldwide economic output of USD 13 trillion by 2030 (Bughin et al 2018).

…”