Dynamic governance for agentic AI: synthesizing dual-use regimes through paradox theory
link.springer.com
Agentic AI can both protect and destroy — and this paper argues only paradox theory can build governance frameworks flexible enough to hold that tension without snapping.
Paradox TheoryDual-Use DilemmaPrincipal-Agent ProblemRegulatory Theory
Theory Briefing
- Agentic AI systems can autonomously act as both beneficial tools and dangerous weapons, creating a dual-use dilemma that static regulations cannot resolve.
- The paper applies paradox theory to governance, arguing that embracing inherent tensions — rather than eliminating them — is key to regulating AI effectively.
- A dynamic, paradox-aware framework is proposed that synthesizes lessons from existing dual-use regimes like biosecurity and nuclear governance.