A Question Bank to Assess AI Inclusivity: Mapping out the Journey from Diversity Errors to Inclusion Excellence

A comprehensive assessment tool is now available: The paper introduces a 253-question framework organized around five key areas (Humans, Data, Process, System, and Governance) that organizations can use to evaluate whether their AI systems are inclusive and equitable—filling a gap that existing risk assessment frameworks have overlooked.

Inclusivity requires systematic evaluation across the entire AI lifecycle: Rather than treating diversity and inclusion as an afterthought, the framework emphasizes that D&I principles must be integrated throughout AI development, from initial data collection and team composition through governance and decision-making structures.

The tool was tested across diverse roles and domains: The question bank was validated through a simulated study with 70 AI-generated personas representing different job roles, suggesting the framework's relevance spans multiple industries and use cases, not just a narrow set of applications.

Standardized measurement of inclusivity is essential for accountability: By providing a structured set of questions rather than vague principles, this tool enables organizations to move beyond general commitments to D&I and instead conduct concrete, comparable assessments of their AI systems' actual inclusivity performance.

Inclusivity is foundational to responsible AI: The paper reinforces that ensuring diverse perspectives and equitable outcomes in AI systems isn't separate from responsible AI—it's central to it, requiring intentional design choices across human teams, data practices, development processes, technical systems, and organizational governance.