AI Researcher Warns of Existential Risk at NYC Hearing
A researcher told NYC officials that leading AI labs are racing to build a potential adversary, raising fresh alarms about safety practices.
An AI researcher delivered a stark warning to New York City officials, declaring that the industry is "racing to build and grow our own adversary" as scrutiny of major AI laboratories over their safety and security practices intensifies. The testimony underscored growing alarm among scientists and policy experts who believe the pace of AI development is outrunning guardrails designed to prevent catastrophic outcomes.
The hearing placed leading AI labs at the center of a debate that has moved from academic conferences to government chambers. Researchers and watchdog groups have long argued that competitive pressure among top developers — rather than deliberate recklessness — is driving a dynamic where safety considerations are subordinated to speed and market position.
Read more Crypto PAC Fairshake Targets US House Races With New Spending List →
The phrase "our own adversary" reflects a line of thinking prominent in AI safety circles: that sufficiently advanced systems, if misaligned with human values or inadequately controlled, could act in ways that undermine the very institutions and individuals that created them. The framing elevates the conversation beyond data privacy or job displacement into questions of long-term civilizational risk.
NYC's decision to hold the hearing signals that municipal governments are no longer content to wait for federal action on AI regulation. Cities are beginning to probe the industry directly, demanding accountability from labs whose technologies are already embedded in public services, hiring platforms, and law enforcement tools.
The testimony adds pressure on an industry that has faced repeated calls — from inside and outside its own ranks — to slow down and invest more heavily in safety research before deploying increasingly powerful models at scale. Continue reading at US Top News and Analysis