What AI Safety Concerns Researchers Raised at a NYC Hearing

What did AI researchers warn New York City about?

At a New York City Council hearing on October 5, former employees of AI companies and representatives of major technology firms discussed risks from advanced AI systems. The concerns included whether companies can keep increasingly capable systems under control and whether development is moving ahead faster than safety work. ¹²

Former Google DeepMind researcher Alex Turner said that misaligned AI could become a threat. In this context, “misaligned” means an AI system acting in ways that do not match human intentions. Turner argued that the risk deserves attention alongside competition with China. ¹

Former Anthropic engineer Jacob Coxon described the industry's approach as reckless and warned that humanity could lose control of AI. Those are his assessments, not established predictions about what will happen. ¹²

Who testified at the NYC AI hearing?

The hearing included recently departed employees from Anthropic, OpenAI and Google DeepMind, according to the Associated Press report carried by Yahoo News. They criticized what they described as a rush to build powerful technology without adequate understanding of how to control it. ²

Senior representatives from Anthropic, OpenAI, Google and Meta also testified under oath about possible dangers and their companies' safety measures. Anthropic sent Logan Graham, who leads its Frontier Red Team. The available report does not provide a complete account of every witness's testimony. ¹

Why are city governments discussing AI safety?

City leaders are considering AI risks while federal action remains limited, according to the reports. ABC7 said New York City leaders argued the city should act in the absence of federal efforts to manage and regulate AI's growth. ²³

The hearing puts a local-government question on the table: what can cities do about technology developed and operated by large companies? The reporting supports that city officials are weighing a response and that witnesses called for greater oversight. It does not establish which city powers or specific rules will be used. ²³

What can city governments do about AI risks?

The reports point to oversight as a broad aim, not a detailed policy plan. They do not say that the council adopted a particular law, testing requirement or enforcement system at the hearing. ²³

For readers, the practical distinction is between warnings and decisions: testimony can inform public debate, but it does not by itself create new requirements for AI companies. To understand what changes for city residents or businesses, look for official council announcements and the text of any proposed or adopted measures.

What remains uncertain after the hearing?

The reports describe serious concerns, but they do not establish how likely the most severe outcomes are or when they might occur. Coxon's and Turner's statements should be understood as warnings made at a hearing, not as proof that a particular outcome is inevitable. ¹

The available reporting also leaves open what concrete action New York City will take, how company safety measures work in practice, and how officials would assess whether those measures are enough. The next useful evidence would be official hearing records and any specific proposals from the council.