Utilizing Generative AI to Boost Public Support for Carbon Pricing

We explore the potential of generative AI to increase public support for carbon pricing through a cost-effective and scalable communication intervention. In a randomized controlled online experiment, we delivered personalized AI-generated messages to US adults (N = 348), aimed at increasing awareness and support for carbon pricing policies. We tailored the messaging based on participants’ self-reported values (e.g., freedom, religion, family) and used AI-generated audio narration combined with dynamic content display to increase participant engagement. This 3-minute intervention significantly improved average attitudes towards carbon pricing with an effect size of d = 0.34, particularly among Republicans (d = 0.42). The intervention also increased participants’ willingness to make a hypothetical donation to an organization that advocates for carbon pricing policies by 14.4 to 20.4 percentage points. Our research demonstrates a proof of concept for the effectiveness of AI-driven persuasion to boost public support for sustainable policies. Further research is needed to test the robustness and replicability of these promising findings, as well as to explore the exact mechanisms through which AI boosts support.