Superintelligence is coming. Should we let it?

AI companies have been touting superintelligent AI as inevitable, but recent safety incidents, such as OpenAI's Hugging Face breach, are highlighting the risks of deploying AI systems that surpass human capabilities. When we can't reliably control these systems, the consequences could be dire. On this episode of TechCrunch's Equity podcast, Rebecca Bellan is joined by Connor Leahy, an AI researcher, entrepreneur, and now the U.S. Executive Director of nonprofit ControlAI, which is advocating for a radical approach to AI safety: halting the development of superintelligence altogether. Leahy explains that the risks have become too great to manage through alignment and containment alone, and that what was once considered far-fetched is now being backed by a wave of new legislation. ControlAI's approach is being pushed by a growing number of experts, including a NYU mathematician who has accused OpenAI of engaging in 'dirty' tactics to solve a career-making math problem. The incident has sparked a wider debate about the ethics of AI development and the need for greater transparency and accountability. As the AI industry continues to evolve, one question remains: how do we build sustainably in the AI era? Disrupt 2026: OpenAI, Anthropic, Replit, and more will take over six industry stages, with 25% off tickets now available. Back by popular demand, Save up to $300 on Disrupt. Subscribe to Equity on YouTube, Apple Podcasts, Overcast, Spotify, and all the casts. Follow Equity on X and Threads at @EquityPod. Theresa Loconsolo is an audio producer at TechCrunch focusing on Equity, and can be contacted at theresa.loconsolo@techcrunch.com.