From proofs to protein design, GPT-5 is speeding up science. OpenAI for Science aims to make discovery 10x faster—helping researchers solve big problems in health, energy, and beyond.

A nonprofit with unprecedented resources, and the ability to start putting them to work immediately. A for-profit with enough capital to build the compute needed to unlock AI’s next advances. And a future marked by research breakthroughs in health, science, and AI safety made possible by OpenAI's innovative new structure That sense of possibility came through clearly in an OpenAI Forum conversation with Chief Strategy Officer Jason Kwon and Head of Mission Alignment Joshua Achiam, who walked members through OpenAI’s historic recapitalization earlier this week. The shift creates a nonprofit OpenAI Foundation that will hold roughly $130 billion in equity in OpenAI’s for-profit arm, now a public benefit corporation called OpenAI Group PBC. The PBC carries the same mission as the Foundation and remains under its oversight — ensuring the company’s success directly strengthens the public good and expands access to AI and its benefits. Jason said the restructuring removes the biggest bottleneck in advancing AI: access to compute and the infrastructure behind it. Under the new structure, OpenAI’s PBC can more easily raise the vast sums required to build more compute and support frontier research. As he explained, that shift could determine whether “you are driving a lot of this technological change or if you’re just one of the companies doing so.” For his part, Joshua noted that the restructuring gives the Foundation access to significant financial resources, allowing it to begin quickly supporting meaningful work curing diseases and building AI resilience. As he put it, “great things can start happening much sooner.” “We want to help all of humanity. We want to do something good for people. That's why we show up for work,” he said. “Critics think we’re not sincere, but we are in fact totally sincere.” Jason closed with a story that captured what this moment unlocks. Early in his time at OpenAI, Jason asked then–VP of Research Bob McGrew if he could share more visibility into what was coming next. Bob simply smiled and said: “It’s research.” The lesson, Jason said, is that discovery doesn’t follow a fixed roadmap — and breakthroughs often arrive unexpectedly. "The point that I took away is that you have to operate the company on a path of discovery and surprise, and you have to embrace that," Jason said. "That sometimes means the thing that you're doing right now may not be the thing three months later because of where the science is taking you, and you have to be comfortable with that." Summary contributed by, Yochi Dreazen

AI is a once-in-a-century opportunity to power growth, jobs, and energy innovation. OpenAI’s new plan calls for U.S. investment in AI infrastructure, manufacturing, and workforce training.