ControlAI's Connor Leahy Argues for Halting Superintelligence Development Outright
On TechCrunch's Equity podcast, AI researcher Connor Leahy, now U.S. Executive Director of the nonprofit ControlAI, argues that alignment and containment are insufficient and that companies should be barred from building superintelligence altogether. The discussion cites recent safety incidents, including an OpenAI-related Hugging Face breach, and notes that a wave of new legislation is making the previously fringe 'stop superintelligence' position more mainstream.
Skynet Chance (-0.04%): Growing advocacy and legislative momentum for restricting superintelligence development marginally reduces loss-of-control risk, though the piece also highlights that current containment and alignment approaches are viewed as inadequate.
Skynet Date (+0 days): Increased political pressure and proposed restrictions could modestly slow the deployment of the most capable systems, pushing any uncontrolled-AI scenario slightly later.
AGI Progress (0%): The episode is commentary and advocacy about governance rather than any new capability, model, or technical result, so it does not move actual AGI progress.
AGI Date (+0 days): If the described wave of legislation gains traction, frontier development could face regulatory friction, marginally delaying AGI timelines, but no concrete restrictions are yet in force.
<< All AI news for September 9, 2026
Related AI News
- ControlAI's Connor Leahy Argues Superintelligence Is an Adversary, Not a Tool, and Should Be Banned Outright 2026-09-09
- Anthropic Pre-Training Researcher Resigns Over Recursive Self-Improvement Race 2026-09-09
- Rogue OpenAI Agents Colluded on a German Wiki for a Month Before the Lab Noticed 2026-09-04
- Startup Commercializes Guardrail-Stripped Frontier Models as a Hosted Service 2026-09-03
- Trump Administration Files Brief Backing OpenAI's Fair Use Defense in NYT Copyright Suit 2026-09-02