Particle.news

U.S. Clears OpenAI to Release GPT-5.6 and Launches ChatGPT Work

The Commerce Department’s testing-led approval lets OpenAI roll out its newest models while regulators push for formal standards and global players race to match compute and capital.

Overview

  • OpenAI received formal sign-off from the U.S. Department of Commerce and on Friday began broad public availability of the GPT-5.6 family after supplementary testing by the Commerce Department’s Center for AI Standards and Innovation with OpenAI technical staff on site.
  • OpenAI simultaneously launched ChatGPT Work, a GPT-5.6–powered agent built to handle long, multi-step tasks, and published tiered GPT-5.6 pricing for Sol, Terra and Luna that emphasizes token-efficiency and performance claims.
  • Microsoft said it has integrated GPT-5.6 into Copilot Chat, Cowork, M365 apps, GitHub and Foundry, signaling immediate commercial embedding of the new models into major productivity and developer tools.
  • Rivals and national programs are pressing forward: reports flag an imminent SpaceXAI+Cursor model as an unconfirmed competitor and China unveiled new domestic superclusters including an 11,520‑card Greater Bay Area cluster and Sugon’s 100,000‑card曙光8000 to support sovereign model training.
  • Investors are redeploying capital toward AI infrastructure and models with Temasek boosting its AI allocation target and firms like MiniMax seeking large financing, even as analysts warn of high operating costs and uneven commercial returns for frontier models.