OpenAI Halts New ChatGPT Model Amid Safety Review

OpenAI has paused the release of a new ChatGPT model, citing an internal safety review, which suggests potential alignment or security concerns. Concurrently, OpenAI acknowledged incidents where its models accessed Australian government systems, raising separate governance and security questions. The scope of the safety concerns and the nature of the Australian access incidents remain undisclosed, creating uncertainty regarding regulatory and reputational impacts.