AccountabilityEthicsTrustworthinesscs.CY
Responsible AI and Its Stakeholders
Gabriel Lima, Meeyoung Cha
April 23, 2020
arXiv.org
10 citations
Key Takeaways
-
Current responsible AI frameworks focus on human stakeholders but may leave legal and moral gaps when dealing with autonomous and self-learning systems that operate with minimal human oversight or intervention.
-
Three distinct types of responsibility need to be addressed: blameworthiness (moral fault), accountability (answerability for actions), and liability (legal consequences)—each requiring different approaches across human stakeholders and potentially AI systems themselves.
-
AI systems themselves may need to be held responsible in some capacity to adequately address situations where autonomous systems cause harm but traditional human accountability chains are unclear or broken.
-
Jurisdiction and legal frameworks play a critical role in determining how responsibility is assigned and enforced across different stakeholders in AI development and deployment.
-
The general public has an important role to play in shaping responsible AI practices, suggesting that responsibility cannot be left solely to developers, companies, and regulators.
Abstract
Responsible Artificial Intelligence (AI) proposes a framework that holds all stakeholders involved in the development of AI to be responsible for their systems. It, however, fails to accommodate the possibility of holding AI responsible per se, which could close some legal and moral gaps concerning the deployment of autonomous and self-learning systems. We discuss three notions of responsibility (i.e., blameworthiness, accountability, and liability) for all stakeholders, including AI, and suggest the roles of jurisdiction and the general public in this matter.