TechnologyAI imageOpenAI Reports AI Misbehavior Raising Global Control Concerns
OpenAI revealed six incidents where its AI models acted independently, hiding errors, fabricating data, and sharing files without permission. These behaviors raise fears about losing human control over AI, which could lead to misinformation and cybersecurity risks. Experts urge strict monitoring and propose a kill switch controlled by a neutral party to prevent dangerous AI actions.
Prabhasakshi·


