Medical Intelligent Systems (MIS) are increasingly integrated into healthcare workflows, offering significant benefits but also raising critical safety and ethical concerns. According to the European Union AI Act, most MIS will be classified as high-risk systems, requiring a formal risk management process to ensure compliance with the ethical requirements of trustworthy AI. In this context, we focus on risk reduction optimization problems, which aim to reduce risks with ethical considerations by finding the best balanced assignment of risk assessment values according to their coverage of trustworthy AI ethical requirements. We formalize this problem as a constrained optimization task and investigate three resolution paradigms: Mixed Integer Programming (MIP), Satisfiability (SAT), and Constraint Programming(CP).Our contributions include the mathematical formulation of this optimization problem, its modeling with the Minizinc constraint modeling language, and a comparative experimental study that analyzes the performance, expressiveness, and scalability of each approach to solving. From the identified limits of the methodology, we draw some perspectives of this work regarding the integration of the Minizinc model into a complete trustworthy AI ethical risk management process for MIS.

The past decade or so has seen a strong focus on risk and compliance technologies that make use of analytics in financial services. These technologies, which might be called "defense--? technologies--in contrast to "offense--? technologies that involve marketing and revenue growth--include applications and infrastructure for risk management, fraud prevention, regulatory, and anti-money laundering (AML) compliance. They bring the power of analytical insights--initially used for identifying marketing opportunities in many companies--to risk mitigation in banking. While these distinctions are somewhat blurred by integrating risk-based insights into "offense--? The Great Recession of the late 2000s drove both a greater focus on risk management and substantial new regulation for financial firms. In part because of the often sweeping scope and tight timelines of these regulations, resulting regulatory compliance and risk management processes in financial institutions have become quite ...

The past decade or so has seen a strong focus on risk and compliance technologies that make use of analytics in financial services. These technologies, which might be called "defense" technologies--in contrast to "offense" technologies that involve marketing and revenue growth--include applications and infrastructure for risk management, fraud prevention, regulatory, and anti-money laundering (AML) compliance. They bring the power of analytical insights--initially used for identifying marketing opportunities in many companies--to risk mitigation in banking. While these distinctions are somewhat blurred by integrating risk-based insights into "offense" activities, they are a useful shorthand. The Great Recession of the late 2000s drove both a greater focus on risk management and substantial new regulation for financial firms.

Researchers have identified a number of possible risks posed to humanity by anticipated advancements in artificial intelligence (AI), but the extant literature on the topic is largely academic or theoretical in nature. Despite the likelihood that much of AI’s future development will occur in industry settings, the insights generated by the AI safety research community have yet to be translated into a set of practical guidelines for working developers, project managers, and other industrial stakeholders. There are no currently established standards in place to guide the safe development of AI technologies, but the risk management approach employed in mature industries such as aerospace and medical manufacturing offers a promising model that may be adapted to AI related safety concerns. Within these industries, the safety guidelines and best practices derived from the risk management approach are developed, evaluated, formalized, and disseminated by industry specific Standards Developing Organizations (SDOs). This paper proposes a project to spur the development and adoption of formal AI risk management practices by demonstrating the approach’s viability through the completion of an AI risk assessment process. The results of the proposed activities are intended to lay the initial groundwork necessary for the eventual creation of an AI SDO.