Trade secrets versus the AI explainability principle
Year: 2022
Type: book-chapter
Abstract: Trade secrets when applied to AI modules have a potential to conflict with the AI explainability principle, which requires AI modules to be explainable and interpretable. Relying on judicial administration and healthcare as two case studies, this chapter first investigates the importance of trade secrets in the AI industry. It then examines when and who needs AI modules to be explainable, and what techniques are currently available to achieve AI explainability. After determining the likely scope of the AI explainability principle, the chapter further examines in which situations the conflict between trade secrets and the AI explainability principle are likely to occur and considers possible solutions to this conflict.
Source: Edward Elgar Publishing eBooks
Authors: Rita Matulionyte, Tatiana Aranovich
Institution: Macquarie University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W4311998398