Making Transparency Advocates: An Educational Approach Towards Better Algorithmic Transparency in Practice

Education can bridge the research-to-practice gap in AI transparency: Despite a decade of Explainable AI research, organizations have struggled to implement algorithmic transparency in practice. The paper demonstrates that targeted educational workshops can motivate employees to become "transparency advocates" who drive cultural change from within their organizations.

Advocacy takes multiple forms and levels: The research reveals that algorithmic transparency advocacy is not a single behavior but exists on a spectrum of different levels, suggesting that organizations should recognize and support various forms of employee engagement with transparency—from informal conversations to formal policy recommendations.

Professional context shapes willingness to advocate: The paper found significant differences across industries, with news and media professionals showing greater likelihood to advocate for algorithmic transparency compared to employees at technology start-ups, indicating that organizational culture and industry norms substantially influence whether individuals will speak up about AI risks.

Hands-on learning drives real-world action: Workshop participants didn't just gain knowledge—they applied it concretely in their workplaces, such as raising algorithmic transparency concerns at organization-wide AI strategy meetings, demonstrating that practical education can translate into measurable organizational change.

Internal advocates may be key to responsible AI implementation: Rather than relying solely on external regulation or top-down mandates, empowering motivated individuals within organizations to champion transparency could be an effective strategy for improving how companies actually deploy AI systems.