Science Communications for Explainable Artificial Intelligence

XAI methods alone aren't solving the communication problem: While explainable AI techniques have been developed to increase transparency, user evaluation studies show that the numerical explanations they typically provide are often ineffective for many users, indicating that technical solutions require complementary communication strategies.

Science communication frameworks offer practical solutions: The paper adapts established communication models from science communications to help AI practitioners design explanations that work better for diverse audiences, suggesting that lessons from how scientists communicate complex topics to the public can improve AI explainability.

One-size-fits-all explanations don't work: The framework emphasizes the importance of understanding and adapting to different user contexts and needs, implying that effective AI explanations must be tailored to specific audiences rather than relying on standardized numerical outputs.

AI literacy in the public requires intentional communication design: The paper highlights that communicating about AI to general audiences requires deliberate application of communication principles, not just technical transparency—suggesting organizations need dedicated attention to how they explain AI systems to non-expert users.

Audience context matters as much as the explanation itself: By integrating communication science principles, the framework recognizes that understanding who users are, what they need, and their existing knowledge shapes whether an AI explanation will actually be understood and trusted.