Runtime Monitoring and Enforcement of Conditional Fairness in Generative AIs
It addresses fairness issues in generative AI deployment, which is an incremental improvement focusing on context-specific enforcement.
The paper tackles fairness concerns in generative AI by defining and enforcing conditional fairness tailored to context, such as demographic fairness in image generation, and develops a prompt injection scheme to enforce it with minimal intervention, validated on state-of-the-art systems.
The deployment of generative AI (GenAI) models raises significant fairness concerns, addressed in this paper through novel characterization and enforcement techniques specific to GenAI. Unlike standard AI performing specific tasks, GenAI's broad functionality requires ``conditional fairness'' tailored to the context being generated, such as demographic fairness in generating images of poor people versus successful business leaders. We define two fairness levels: the first evaluates fairness in generated outputs, independent of prompts and models; the second assesses inherent fairness with neutral prompts. Given the complexity of GenAI and challenges in fairness specifications, we focus on bounding the worst case, considering a GenAI system unfair if the distance between appearances of a specific group exceeds preset thresholds. We also explore combinatorial testing for assessing relative completeness in intersectional fairness. By bounding the worst case, we develop a prompt injection scheme within an agent-based framework to enforce conditional fairness with minimal intervention, validated on state-of-the-art GenAI systems.