BREAKING: Florida seeks injunction against OpenAI

Substack · Watch/Read · (gathered 2026-09-30)

• Gary Marcus argues that Florida’s call for an injunction against OpenAI’s ChatGPT development, citing “tens of thousands of security incidents,” is the “bigger news” and a necessary step that other states and countries should follow.
• He expresses skepticism about Nvidia’s newly announced “AI Agent Safety” platform (OpenShell and Sentry), viewing it as an implicit acknowledgment of risk but potentially “software for PR” like Nemoclaw, and reiterates that “LLMs in themselves should never be trusted.”
• Marcus warns that without immediate injunctions against OpenAI, society is “playing Russian roulette” with the risk of “massive and possibly deadly AI-induced infrastructure problems,” as OpenAI cannot properly regulate its own technology.