The Stories We Govern By: AI, Risk, and the Power of Imaginaries

Competing narratives about AI risk are shaping real governance decisions. The paper identifies three dominant story groups—existential risk proponents, accelerationists, and critical AI scholars—each with fundamentally different views on what AI dangers matter most. These aren't just academic debates; they actively influence which regulatory approaches get adopted and which get ignored.

Each narrative group has blind spots rooted in different assumptions about the future. Existential risk advocates focus on hypothetical catastrophic scenarios, accelerationists emphasize transformative potential, and critical scholars highlight present-day harms from inequality. The paper shows these groups disagree not just on solutions, but on basic diagnoses of what problems exist, potentially leaving important issues unaddressed depending on which narrative dominates policy.

Current AI governance may be narrowing rather than broadening our options. The authors warn that when one imaginative framework becomes dominant in policy-making, it can actually restrict the space for alternative governance approaches. This means the story we choose to believe about AI risk could inadvertently lock us into less effective regulatory strategies.

We need governance grounded in pragmatism rather than speculative certainty. The paper argues against what it calls "speculative dogmatism"—the tendency to treat any single narrative about AI's future as definitive truth. Instead, responsible governance should acknowledge uncertainty and remain flexible enough to address multiple types of risks simultaneously.