OpenAI is preparing to launch an open-weight AI model as soon as next week, described as similar to the o3 mini model and equipped with reasoning capabilities. Meanwhile, JPMorgan has expanded its research coverage to include private companies, beginning with OpenAI. The bank issued a 24-page initiation report outlining its strategic thesis on OpenAI's mission to ensure artificial general intelligence benefits humanity, while also cautioning about the company's aggressive spending on talent and research and development. In parallel, the open-source AI community is making strides with the release of GLM-4.5 by Zhipu AI (now known as Zai_org), a model that is competitive with leading closed-source models like Claude 4 Opus and Gemini 2.5 Pro across multiple benchmarks. GLM-4.5, licensed under MIT and backed by $1.5 billion in funding from Tencent and others, offers reasoning, coding, and agentic capabilities and is available through Novita AI's API. Additionally, the Qwen3 open-source model from AI-Labs-China has outperformed GPT-4 and Gemini in logic, math, and coding exams, signaling a growing challenge to closed-source AI dominance. Concerns have also been raised about OpenAI deploying an AI classified as high risk for biological capabilities, with some experts questioning whether the company has adequate tools to manage such risks.