Dominic Vincent Ligot (CirroLytix Research Services) has posted “Generative AI Safety: A Layered Framework for Ensuring Responsible AI Development and Deployment” on SSRN. Here is the abstract:
This article discusses a multi-layered framework to address the challenges around Generative AI safety. Structuring GenAI safety across key stages: Training Data, Algorithm, Inference, Publication, and Societal Impact, each layer integrates principles of quality, fairness, transparency, accountability, and human oversight, enabling targeted interventions throughout the AI lifecycle. By incorporating distinct values and intervention strategies at each stage, this framework addresses the unique risks associated with GenAI while providing a structured pathway for responsible development. The article also discusses practical implications for applying the framework in the Philippine context, where AI readiness shows both high interest and structural challenges, and explores its alignment with current legislative initiatives. This framework is intended to serve as a guide for researchers, practitioners, and policymakers to promote responsible GenAI development, advocating a future where AI innovation is balanced with ethical responsibility and societal well-being.