Using Legends Into AI-Based Business Decision Making: Embedding Ethics, Cybersecurity and Resilience

Navigating the complex landscape of sustainability regulations-encompassing green finance, ethics, human rights, and social implications-presents a significant challenge for large companies. Driven by cost optimization and profit maximization, these organizations often overlook crucial social costs in their decision-making. While generative AI offers powerful tools for regulatory implementation, ensuring that social aspects are not neglected requires careful consideration. This article explores innovative methods for embedding ethical AI directly into optimization models from the outset, eliminating costly delays and reactive adjustments. Using examples from sustainable finance and gender equality, we demonstrate a novel approach to embedding regulations by generating “champion” positive examples that inform decision-making. We further emphasize that secure and resilient digital infrastructures are essential for ensuring the reliability and integrity of these AI-driven systems, especially when they influence sensitive decisions in compliance-heavy domains. This proactive strategy provides an alternative to constant postdeployment model auditing. Furthermore, we propose maintaining multiple model versions, enabling immediate adjustments to decision-making processes in response to new information and regulations, leading to more efficient and real-time compliance.