ControlAI's Connor Leahy Warns of Superintelligent AI Control Risks

9 Sept 2026, 9:35 pm · 3d ago · 1 min read · TechCrunch AI

On TechCrunch's Equity podcast, AI researcher and ControlAI US Executive Director Connor Leahy discussed growing safety and control concerns surrounding the rapid development of superintelligent artificial intelligence systems. Highlighting recent industry incidents, including OpenAI's Hugging Face security breach, Leahy argued that tech companies are treating superintelligence as an inevitability without establishing reliable safety and containment mechanisms. The discussion focused on the systemic challenges of controlling systems that exceed human capabilities and the urgent necessity for regulatory safeguards and technical boundaries before deploying advanced autonomous models into enterprise and public infrastructure.