📰 Key Takeaways

New York City Mayor Zohran Mamdani and Broad City star and activist Ilana Glazer talked AI on Monday night at Webster Hall in Manhattan, at “The People’s AI Assembly,” an event hosted by the think tank AI Now Institute. Mamdani urged New Yorkers not to accept the fatalism that AI will inevitably take their jobs, mess with their kids’ minds, and erode their dignity. He argued that the rapid rollout of the technology isn’t inevitable, and that public officials should demand accountability from AI companies. He pointed out that the federal government has barely regulated AI at all, so cities have to step up on their own.

On the concrete side, Mamdani ordered a ban on AI in elementary and middle schools last month. Glazer, who’s a parent, admitted that AI coming into classrooms makes her uneasy, since nobody asked for consent. This summer he also proposed a policy requiring landlords, brokers, and real estate platforms to clearly disclose whether listing photos or videos were made with AI, so people don’t get lured in by misleading “bait” listings.

Mamdani criticized society for “recklessly adopting technology without reckoning with its limits,” citing the US military’s airstrike on a school in Iran this February that killed more than 150 people (at least 123 of them children). Pentagon internal investigators told Bloomberg that over-reliance on AI in target identification was one of the causes. He said the industry’s pitch that AI is beyond question and that only its developers can oversee it echoes what Facebook said in 2008, and is really a way to dodge accountability and protect IPO valuations. He also slammed the “complete abandonment of legislation” at the federal level, noting that AI debates are often framed as an arms race with China, which pushes human concerns aside. Anthropic is expected to go public this year, and OpenAI says it will go public in 2027.

💬 JudyAI Lab Take

New York City Mayor Mamdani and activist Glazer talked AI at an event hosted by the AI Now Institute, arguing that the rapid rollout of the technology isn’t inevitable. This story matters because there’s almost no regulation at the federal level, and local governments have started setting limits themselves.

For AI builders, this case points to a trend: rules are shifting away from top-down, uniform legislation and toward specific settings like cities, schools, and the housing market, each setting its own. Mamdani first banned AI in elementary and middle schools, and also required that listing photos or videos made with AI be disclosed. Both are requirements aimed at a single use case. The original article also notes that over-reliance on AI in target identification was reportedly a cause of one airstrike, which makes “did you reckon with the technology’s limits?” the thing the public looks at when it examines AI products. If the industry keeps pitching AI as beyond question, outsiders will just lean toward answering with bans. Products that clearly spell out their capability boundaries and who’s responsible will have an easier time with the different requirements popping up in different places.

My suggestion: add an “AI-generated” label to your own AI features, and write a short note about which situations they’re not suited for.

📅 Source Info

🔗 Further Reading