The Current State of Industrial Practice in Artificial Intelligence Ethics

As Artificial Intelligence (AI) systems become increasingly widespread, we have begun to witness various failures highlighting issues in these systems. These incidents have sparked public discussion related to AI ethics and further accelerated the on-going academic discussion in the area. High-level guidelines and tools for managing AI ethics have been introduced to help industry organizations make more ethical AI systems, but we currently know little about the state of industrial practice. Have these guidelines been adopted by the software industry for developing AI solutions? Are these failures that make the news just the tip of the iceberg? We provide insights into the current state of practice by presenting the results of a survey of 211 software companies.

Focus: AI Ethics/Policy

Source: IEEE

Readability: Expert

Type: PDF Article

Open Source: Yes

External URL: https://jyx.jyu.fi/bitstream/handle/123456789/68453/vakkuriym.pdf?sequence=2&isAllowed=y

Keywords: Artificial Intelligence, Ethics, Software Development

Learn Tags: Bias Ethics Fairness Framework

Summary: Using survey data from 211 software companies, this article provides needed insight into the current state of AI ethics in the data industry. For practitioners, the data can also serve as a way to benchmark where an organization stands.