New York City lawmakers questioned representatives of major A.I. companies at a City Council hearing on Monday, but left with little clarity about how the firms planned to address looming threats. Public officials asked for answers; the companies whose systems raised the questions didn't explain key parts of their risk planning.
“I don't know,” said Morgan Dwyer, who runs OpenAI's policy development and operations, when City Council speaker Julie Menin asked how the companies were quantifying “the risk imposed by A.I. in the worst case catastrophic scenario.” Dwyer also said she didn't think it mattered whether the chance of catastrophe was 1 percent, 10 percent or 20 percent.
Questions With No Clear Answers
The hearing produced no quantified estimate of the risk of a cataclysmic event. The companies also didn't say whether they had insurance for such events. Lawmakers asked direct questions about how firms were assessing and preparing for the worst-case outcomes tied to their technology. The answers weren't there.
Lawmakers also asked whether a company could stop an A.I. model that failed an internal safety test or an external review. The companies didn't explain whether they could. So one basic question remained unresolved: what happens when a model fails a safety check, and who can halt it?
These gaps matter to the hearing’s central issue: companies developing A.I. were asked to account for potential harms, but offered few assurances about safeguards. The account of the hearing identifies no plan presented by the companies to quantify catastrophic risk, no insurance answer, and no explanation of a stop mechanism for a model that fails review. Public questioning produced an outline of unanswered concerns, not a clear account of how the firms would respond.
The Standard the Companies Were Asked to Meet
Menin rejected Dwyer’s response. “I'm sorry, but to say you don't know and it doesn't matter is flippant at best,” she replied. Menin also said none of the risk levels Dwyer named was remotely acceptable. Their exchange made the disagreement plain: the speaker demanded that companies address the possible scale of a catastrophe, while the OpenAI representative said she didn't know the probability and didn't consider the precise percentage important.
Dwyer said OpenAI shouldn't train models that can't be safely kept under human control. That statement set out a standard, but the hearing yielded few answers about how the industry would provide safeguards against catastrophic scenarios. No one explained whether a model could be stopped after failing an internal safety test or external review.
The source reports no community-led response, mutual-aid effort, or direct action at the hearing, and it describes no legislative or electoral remedy. It names no nonprofit or other institutional helper, either. What it describes is a public hearing in which lawmakers questioned powerful technology companies and left without clarity on risks, insurance, or control. Officials’ questions went on the record. The companies’ safeguards remained largely unanswered.