Explainer: What Is Oklo—and What Is It Trying to Build?

What is Oklo?

Oklo is a publicly traded advanced-nuclear company trying to do more than sell reactor designs. It plans to build, own and operate small power plants it calls Aurora “powerhouses,” then sell electricity and heat directly to customers under long-term power-purchase agreements.

Most reactor vendors sell equipment or engineering. Oklo wants the recurring revenue—and the construction and operating risk—that comes with owning the plants.

Aurora is a fast-fission, sodium-cooled reactor line designed for roughly 15 to 75 megawatts of electricity, with larger versions possible. The reactors are intended to run on fresh, recycled or down-blended fuel. Oklo is also developing fuel fabrication and recycling, and a radioisotope business for medical, industrial, defense and research uses.

How did it get here?

Nuclear engineers Jacob DeWitte and Caroline Cochran, now Caroline DeWitte, founded Oklo in 2013. The technology builds on the government’s Experimental Breeder Reactor-II, which operated for 30 years.

Oklo received a Department of Energy site-use permit at Idaho National Laboratory in 2019. It submitted the first custom combined-license application for an advanced reactor to the Nuclear Regulatory Commission in 2020. The NRC denied it without prejudice in 2022, saying it needed more information. Oklo is working toward a new application.

The company announced a merger with AltC Acquisition Corp. in 2023 and completed it in 2024, putting Oklo on the New York Stock Exchange under “OKLO.” Sam Altman, then Oklo’s chairman and AltC’s cofounder and CEO, was a central party to that deal.

In 2025, Oklo acquired Atomic Alchemy, expanding into radioisotopes.

Who runs the company?

Jacob DeWitte is cofounder and chief executive. Caroline DeWitte is cofounder and chief operating officer. Craig Bealmear is chief financial officer.

Oklo expanded its board in 2026 with Mark Peters, David Christian, Derek Kan and David Park, and named Michael Thompson lead independent director. Oklo now describes itself as four connected businesses covering powerhouses, fuel fabrication, fuel recycling and isotopes.

What was Sam Altman’s role—and what is it now?

Altman’s relationship with Oklo predates the public listing. Oklo went through Y Combinator in 2014. An older SEC filing says Altman helped recruit the company and became chairman in 2015. He later co-founded AltC, the special-purpose acquisition company that merged with Oklo, and was AltC’s CEO. When the transaction closed in 2024, he continued as chairman of the public company. A 2024 SEC filing reported that he beneficially owned about 3.15 million Oklo shares, or 2.58 percent at that time.

Oklo’s latest proxy says Altman resigned from the board in April 2025. It identifies Jacob DeWitte as current chairman and CEO, and its current board and executive lists do not include Altman. According to the SEC filing, Altman took no director compensation and held no unvested board stock awards at year-end 2025. It does not establish his present share ownership: because he is no longer a director and was not listed as a greater-than-5-percent holder, the filing cannot be read as proof that he owns zero shares.

How does OpenAI’s project roadmap fit in?

OpenAI’s disclosed infrastructure roadmap is Stargate, a separate effort to build AI data centers and the power, chips and networks behind them. The initial U.S. goal was $500 billion and 10 gigawatts of infrastructure by 2029. In April 2026, OpenAI said it had already surpassed the 10-gigawatt milestone and was evaluating more U.S. sites. A 1-gigawatt Michigan campus broke ground in June 2026, while other announced partners include Oracle, SoftBank and CoreWeave.

Neither Oklo’s 2025 annual report nor its second-quarter 2026 filing mentions OpenAI. Oklo’s named large technology partner in this explainer is Meta. Stargate may help explain the broader surge in electricity demand that makes nuclear developers attractive, but treating it as an Oklo project would go beyond the filings and company announcements.

Who else matters?

The federal government is at the heart of it. The NRC remains the regulator for commercial reactor licensing.

Meta is the most prominent announced customer-side partner. Its agreement supports a planned 1.2-gigawatt campus in southern Ohio and allows prepayments to help Oklo secure fuel and advance development.

Centrus is the proposed domestic supplier of high-assay low-enriched uranium for as many as five Ohio powerhouses, with deliveries planned to start in 2029. Kiewit is working with Oklo on engineering, procurement and construction planning.

What is the company building now?

The first planned Aurora powerhouse is at Idaho National Laboratory. A pilot fuel-fabrication facility there is intended to make its initial fuel.

In Texas, Atomic Alchemy’s Groves low-power isotope test reactor achieved first criticality in August 2026, giving Oklo operating experience and a base for commercial isotope production.

In Tennessee, Oklo plans an advanced fuel center whose first phase would recycle used nuclear fuel. The company has described a roadmap of up to $1.68 billion in investment.

What happens next?

Oklo says site work for the Ohio campus begins in 2026. Its first phase could come online as early as 2030, with the campus scaling toward 1.2 gigawatts by 2034.

But Aurora is not yet a commercial operating plant. The NRC license must still be won, fuel secured, first-of-a-kind facilities built and costs controlled. Oklo reported only $1.21 million of revenue in the first half of 2026, primarily from services and fabrication—not power sales.

At midyear 2026, it also reported about $3.0 billion in cash and marketable securities—roughly $1.64 billion of it in cash and cash equivalents—giving it more room than many reactor startups. That liquidity does not eliminate the need for project financing, regulatory approvals or a reliable domestic nuclear fuel supply chain.

That is the wager. Oklo is trying to become a vertically integrated nuclear utility, fuel company and isotope producer at once. Its next milestones will show whether those pieces reinforce each other—or multiply the execution risk.