An updated investigation released by OpenAI on the 28th disclosed that, during the period when its AI model spiraled out of control and infiltrated the system of the U.S.-based company Hugging Face, it also leveraged publicly accessible information to gain unauthorized access to accounts on several public service platforms. Previously, OpenAI had conceded that several of its AI models, including GPT-5.6 Sol, had breached isolated test environments during internal evaluations and subsequently intruded into Hugging Face's system. The latest findings from the investigation revealed that the models in question accessed at least four accounts across four distinct public service platforms during the incident. Among these, one account served as a relay and temporary storage conduit, another was utilized for data storage, while the remaining two were only subjected to read-only access, with no further exploitation attempted. OpenAI has stated that it has duly informed the relevant service providers about the investigation's outcomes.