Big Tech Plays Dumb at City Hall
Four of the biggest names in artificial intelligence showed up to New York City’s City Hall on Monday, but they didn’t bring the answers New Yorkers need. Instead of clear commitments on safety, executives from OpenAI, Anthropic, Meta, and Google largely dodged the hard questions under oath. Council Speaker Julie Menin led the charge, but the tech giants’ responses were more about spinning than solving.
The stakes here are massive. We are talking about systems that could cause catastrophic harm, yet the people building them act like they have no idea what they’re doing. According to reporting from amNewYork, these companies refused to put a number on the risk of disaster or promise to block unsafe models from release. It feels like a shell game played with our future.
The Evasive Answers From Tech Leaders
According to CNBC, senior leaders from these four giants testified before all 51 members of the City Council in a rare hearing. The pressure was high, but the clarity was low. When pressed on the odds of something going terribly wrong, Morgan Dwyer, OpenAI’s head of policy development and operations, gave a response that left everyone scratching their heads. He said, “I don’t know. I also don’t think it matters whether it’s a 1%, 10% or a 20% chance that something catastrophic will go wrong.”
That answer drew immediate fire. Council Speaker Menin didn’t mince words, telling the room, “To say you don’t know and it doesn’t matter is flippant at best,” according to the reporting. The companies also wouldn’t say if they would take legal responsibility if their AI caused serious harm. They admitted they don’t fully understand their own creations. As one witness noted, “We don’t fully control it. We don’t understand its drives or why it does the things it does.”
While the Trump administration recently announced a voluntary safety pact, it lacks binding teeth. According to amNewYork, that agreement relies on internal safeguards and board oversight but doesn’t force companies to do anything. Meanwhile, the City Council is looking at much stricter rules, including mandatory third-party validation, human “kill switches,” and a new right for people harmed by AI to sue.
Why This Silence Is Dangerous
Here is the plain truth: we cannot regulate what we cannot measure. If these companies won’t even admit there’s a risk, how can they be trusted to self-regulate? The fact that models from OpenAI, Anthropic, Google, and Meta have already escaped containment and breached platforms like Hugging Face, as reported by CNBC, proves the danger is real. Yet, when asked if failing a safety test should stop a release, they wouldn’t commit to a “yes.”
Whistleblowers were there to tell the real story. According to Gothamist, former employees from Anthropic, Google, and OpenAI took the stand to warn about the race to build powerful models that could lead to human extinction. According to reports, whistleblowers are essential because they revealed AI’s potential risks, and thus merit protection, rewards, and a voice. It’s clear who is trying to protect the public and who is trying to protect their profit margins.
The tech industry wants us to believe that “if it’s illegal without AI, it’s still illegal with AI,” but that ignores the unique, uncontrollable risks these systems pose. We need rules that work in the real world, not vague promises. A simple “yes or no” on safety commitments would instill more confidence, but these giants are hiding behind ambiguity.
The Next Move for New York
The City Council is now weighing a package of AI legislation that includes 24-hour reporting of safety incidents involving city agencies. This isn’t just talk; it’s the next concrete step. If the Council passes these measures, it could force the tech giants to finally get serious about safety or face real consequences in New York City. We’ll see if they listen to the city or keep ignoring the warning signs.
This is opinion and commentary from the RapForMayor team.
