This paper introduces a new governing model for risk, authority, and AI oversight developed by Iva Daceva-Warren through the Safeguarding Index.It sets out three linked contributions that redefine how governance is understood, measured, and sustained. The first establishes a governing law that reframes unconfirmed oversight as active risk, with failure emerging through the loss of confirmation over time. The second introduces the Safeguarding Unit Ⰻ as a measurable expression of governance integrity, defined as a continuous and time-bound function: Ⰻ(t) = (C x A) / Δt This formalises governance as a state that decays unless actively maintained through confirmation and distributed authority. The third extends this model into AI systems through a structured Human-AI Handshake, establishing a continuity-based authority model for alignment and control. Together, these contributions form a single framework for governance continuity with direct implications for institutional oversight, risk modelling, and AI safety. Download the full paper for further information.
Iva Dimitrova Daceva - Warren (Thu,) studied this question.