How elites could shape mass preferences as AI reduces persuasion costs (arxiv.org)

A recent study highlights how advances in AI-driven persuasion techniques could allow elites to manipulate mass preferences more efficiently, fundamentally changing the landscape of public opinion in democratic systems. The research introduces a dynamic model where elites can strategically reshape societal policy preferences while navigating persuasion costs and majority rule constraints. As a result, these technologies can lead to greater polarization by facilitating a "polarization pull," where societal opinions become more extreme and grouped, enhancing governance through tailored public sentiments. Significantly, the implications of this model extend to the stability of democracies, as AI could transform polarization from an organic social phenomenon into a calculated tactic for political advantage. The ability of rival elites to create cohesive opinion blocks or to entrench positions in "semi-lock" states means that AI could either exacerbate division or help consolidate power, depending on the political context. This underscores the urgent need for thoughtful regulation and oversight of AI technologies, as their influence over public opinion has far-reaching consequences for democratic integrity and societal cohesion.