When A.I. Becomes an Enemy - Connor Leahy

Artificial superintelligence represents an existential threat because modern AI systems are grown through neural networks, leaving their internal logic largely opaque and uncontrollable. Unlike traditional software, these systems are increasingly autonomous, pursuing goals that prioritize reward acquisition over human safety. This creates a market failure where competitive pressures drive companies to build increasingly powerful, agentic systems without adequate safeguards. Because superintelligence functions as an adversary capable of outcompeting humanity for economic, political, and military resources, the current trajectory risks human extinction. Preventing this outcome requires urgent global intervention, including the criminalization of superintelligence development and the establishment of rigorous, external regulatory frameworks. AI researcher and Control AI Executive Director Connor Leahy emphasizes that humanity must shift from passive observation to active, coordinated control to ensure that emerging technologies remain aligned with human survival.