AI giants give few clear answers to key safety questions at NYC Council hearing amid whistleblower warnings

The New York City Council's hearing highlights a persistent challenge in AI governance: obtaining concrete, quantifiable risk assessments from leading developers. While lawmakers sought specific figures on catastrophic risk, the companies did not supply them. This suggests a gap between regulatory ambitions for clear safety metrics and the current willingness or ability of AI firms to provide them.

This lack of specific risk quantification complicates the development of clear regulatory frameworks. Without agreed-upon metrics or a shared understanding of what constitutes a 'catastrophic' risk and its likelihood, regulators are left to craft rules based on more general principles rather than empirically-driven safety targets. This could lead to either overly broad regulations or ones that miss specific, emerging threats.

The dynamic observed in New York is not unique; similar discussions are unfolding globally. The reluctance to commit to specific risk numbers could reflect several factors, including the inherent difficulty in predicting future AI behaviors, concerns about legal liability, or a strategic choice to maintain flexibility during early-stage development. This makes establishing international or even national standards difficult.