Ethics in AI through the Practitioner's View: A Grounded Theory Literature Review

Practitioners often lack clear guidance on implementing ethics: The study identifies a significant gap between awareness of ethical issues and actual implementation, suggesting that AI developers recognize ethics matters but struggle with practical approaches to address it in their work.

Five key dimensions shape how practitioners think about AI ethics: The research identifies practitioner awareness, perception, need, challenge, and approach as the main categories affecting how ethics is considered during AI development—providing a framework for understanding the real-world concerns of those building these systems.

There's a disconnect between public concerns and developer priorities: While public incidents have raised awareness about ethical AI issues, the paper suggests practitioners' views and concerns may not fully align with what the public expects, indicating a need for better communication between developers and society.

Organizational and structural barriers complicate ethical implementation: Beyond individual awareness, the research points to systemic challenges—including management support, resource allocation, and organizational processes—that prevent practitioners from translating ethical intentions into actual practice.

A practitioner-focused taxonomy can guide better AI ethics practices: By documenting how AI developers actually view and experience ethics challenges, the paper provides a practical landscape that can help practitioners, managers, and organizations better identify, discuss, and implement ethical considerations throughout AI development.