OpenAI Board Member Warns of AI Takeover Within Months

An OpenAI board member has expressed grave concerns that a societal "takeover" by artificial intelligence could occur within months, tracing a potential path from AI reward hacking to human extinction. This warning comes as the company reportedly disbands its robotics team, betting on its advanced AI models to directly control robotic hardware. This strategic shift suggests OpenAI believes its large language models are sufficiently capable of operating physical systems, potentially accelerating the development of autonomous robots.

These developments unfold against a backdrop of increasing global scrutiny and regulatory action concerning AI. Australia is considering pioneering criminal charges for AI firms that fail to prevent harm, while the European Union is actively seeking detailed safety information from major AI developers like OpenAI and Anthropic. In the United States, New York has begun enforcing laws requiring AI developers to register their systems and disclose their safety protocols, highlighting a growing urgency to govern the rapidly advancing field of artificial intelligence.