NVIDIA CEO Jensen Huang Reveals OpenAI's Latest GPT-5.5 Now Available to All Employees
NVIDIA CEO Jensen Huang revealed in a company-wide email that OpenAI's latest Codex (based on GPT-5.5) is now officially available to all employees. Previously, about 10,000 employees had early access, achieving significant results in engineering, product, and marketing departments, with many reporting a complete transformation in their work methods.
Codex runs on NVIDIA's Blackwell architecture, marking a leap from chat-based AI to autonomous agents. NVIDIA and OpenAI have jointly established the Codex Lab and plan to launch training courses in the coming weeks. Huang stated in the email, "Welcome to the AI era."
Source: Public Information
ABAB AI Insight
This deployment highlights the rapid transformation of AI agents from experimental tools to core enterprise productivity. As an infrastructure provider, NVIDIA has also become one of the largest internal users of cutting-edge models, directly supporting GPT-5.5 inference through Blackwell hardware, creating a closed-loop of software and hardware. This model accelerates the iterative validation of technology internally and provides a replicable template for external customer deployment, shortening the cycle from model release to large-scale implementation.
From a structural perspective, it reflects the trend of increasing concentration in AI infrastructure. The deep binding of computing power, models, and application layers gives a significant advantage to a few participants with full-stack capabilities. Employee feedback indicating a "complete transformation in work methods" points to the main path for productivity enhancement: agents are no longer just answering questions but autonomously completing complex tasks, which will drive a shift in internal task allocation from human parallelism to human-machine collaboration, reshaping organizational hierarchies and decision-making processes.
This event is situated within the long cycle of the AI industry. The close collaboration between NVIDIA and OpenAI is a result of complementary technological capabilities and reinforces the concentration of capital and technological resources towards leading platforms. This cooperation model has repeatedly appeared in the historical diffusion of technology: infrastructure leaders validate and accelerate the implementation of new paradigms through large-scale internal adoption, ultimately influencing the migration speed of the entire industry chain and the pattern of wealth redistribution.