Former AI researchers are sounding the alarm over how quickly advanced AI systems are developing. At a New York City Council hearing, researchers discussed concerns around AI control, model alignment and gaps in current safety measures, while industry representatives addressed the challenges of assessing catastrophic risks.