AI-Induced Catastrophe

The unregulated, breakneck advancement of artificial intelligence—spanning autonomous systems, general-purpose models, and deep integration into critical infrastructure—creates a convergence of accidental and intentional existential risks. AI systems are already being deployed in cyber warfare, autonomous weapons, and global financial networks, while robust alignment, safety verification, and international governance lag dangerously behind. A self-improving AI or a coordinated AI-driven cyberattack could trigger cascading failures: collapse of power grids, financial systems, and communication networks, or unintended escalation of armed conflict between nuclear powers, all unfolding faster than human decision-makers can respond. Unlike climate change, which operates on decadal timescales, AI risk can manifest suddenly and irreversibly, making it the most probable near-term pathway to global catastrophe.