Ethical AI and Librarianship
A Resource Guide
Ethical Application of Artificial Intelligence Framework
Field
|
Description | |
---|---|---|
Title | Ethical Application of Artificial Intelligence Framework | |
Type | Reports | |
Creator | ACT-IAC (American Council for Technology-Industry Advisory Council) | |
Link | ||
Creation Date | 10/08/2020 | |
Last Updated Date | -- | |
Summary | Developed by Artificial Intelligence Working Group under ACT-IAC Emerging Technology Community of Interest, this report introduces the Ethical Application of Artificial Intelligence Framework (EAAI), a tool designed to assess the ethical use of AI, with a particular focus on U.S. government contexts. The framework presents an index-based evaluation method (EAAI Index) that measures AI systems across five key ethical components: Bias, Fairness, Transparency, Responsibility, and Interpretability. Each component is defined by specific indicators, implications, and strategies for monitoring and measurement. 1) Bias: refers to the influences that affect decisions of the people and organizations involved in AI, such as diversity of teams involved in the development and deployment of AI systems across nine federally protected classes (e.g., sex, race, age. 2) Fairness: involves the inclusive and impartial treatment of individuals and groups to ensure equitable outcomes from AI systems. 3) Transparency: focuses on the AI system’s ability to clearly explain how and why it produces its results. 4) Responsibility: emphasizes accountability of people and organizations for how AI is used and the consequences of its outcomes. 5) Interpretability: ensures AI outputs are consistent, meaningful, and understandable to a wide range of stakeholders. The framework concludes with a scorecard that calculates an EAAI Index score ranging from 5 to 250, which enables practitioners to quantify the ethical strengths and risks of AI systems. | |
Topic | Ethical AI. | |
Source and Link | ACT-IAC (American Council for Technology-Industry Advisory Council). www.actiac.org | |
Access | Open | |
Accessibility | -- | |
Audience | General. | |
Platform or Format | Document (.pdf) | |
Length | 32 pages | |
Geography | USA | |
Language | ENG | |
Description Date | 06/18/2025 |