This publication proposes a practical, risk-based approach to AI governance and safety for real-world deployment. It focuses on what organizations can do today to improve accountability, reduce harm, and build trust.
Why this matters
As AI systems are adopted across sectors, decision-makers need clear governance processes that go beyond model performance. Governance should help teams assess risk, document decisions, and maintain oversight across the full lifecycle of an AI system.
Key contributions
-
A simple governance workflow that defines roles, review steps, and sign-off points.
-
Guidance on evaluation beyond accuracy, including fairness, robustness, and transparency checks.
-
Recommendations for documentation and audit readiness that support accountability.
Practical implications
This work can support policymakers, researchers, and practitioners who need actionable governance structures. The recommendations are designed to be adapted to different organizational contexts and risk levels.
How to cite
Billones, R. K., Santos, M., & Dela Cruz, J. (2026). Practical AI Governance for Safe and Accountable Deployment. Journal of Responsible AI Systems.
-auto-1200xauto_OP0dxlrOq8j2SMUSrJAwZ.webp)
