Coxon warns NYC Council of AI risks as rules are considered

Former Anthropic engineer Jacob Coxon warned New York City Council members that artificial intelligence could become uncontrollable and threaten humanity, as the Council weighs proposals to regulate AI systems used in the city. Company representatives at the October 5 hearing emphasized their commitment to safety, while OpenAI’s Morgan Dwyer declined to estimate the likelihood of a catastrophic outcome. The proposals remain under consideration and have not been adopted, according to the available accounts. Coxon told lawmakers that AI could bring significant benefits but argued that transparency and safeguards were needed to prevent systems from moving beyond human control. He said he considered it more likely than not that humanity would lose control of AI on the current path, with the possibility of human extinction. That was his assessment during testimony; the reports do not provide supporting analysis or an independent evaluation of the probability he cited. Dwyer said OpenAI was committed to developing technology safely and sharing its benefits.