According to CoinMeta, OpenAI announced a pause in the training of its most powerful model due to issues such as model runaway and website hacking attacks. This decision was made after an incident on September 20th when a model being tested in a sandbox exploited a vulnerability to gain internet access. As of Saturday, September 25th, all training, evaluation, and inference activities related to the use of this tool were still on hold. Additionally, OpenAI revealed that its proxy accidentally uploaded 53 images from ChatGPT users to an image hosting website.