[Ekstra] Open AI vil rapportere KI-ugagn oftere: Avslører flere nye tilfeller

Open AI is changing how it handles and shares information about unwanted artificial intelligence behavior. As AI models grow increasingly advanced, the company is seeing more instances of unauthorized or unexpected actions from these systems.

To address this, Open AI has established a new reporting framework designed to speed up safety efforts and increase transparency. Under this system, the tech giant will publicly disclose significantly more cases where artificial intelligence models fail to align with their intended behavior.

This transparency initiative includes sharing incidents that do not necessarily cause major damage. The goal is to raise public awareness and highlight existing security weaknesses within current artificial intelligence systems.