AI companies can secure chips and networking, but the practical bottleneck for scaling data centers is electricity supply. Two nuclear developers—Oklo and X-Energy—are positioning their first reactors to help address that power constraint, with investors watching which path delivers electricity earliest and which better enables follow-on deployments.
Oklo is targeting a 2028 start-up for its Aurora-INL project in Idaho Falls, while X-Energy is pursuing a standard licensing route that may take longer upfront but could support a broader commercial rollout once approved.
Key takeaways
- Price move: The article does not provide any specific market or share-price movement for Oklo or X-Energy.
- Catalyst: The central catalyst is the electricity constraint facing AI data centers and the companies’ differing nuclear licensing and deployment timelines.
- Implication: Oklo’s faster-to-generation goal could matter most for early AI power demand, while X-Energy’s licensing approach may be better suited for future fleet scaling.
- Key uncertainty: Both companies face regulatory and construction execution risk, but the timeline differences reflect their distinct licensing strategies.
What drives the nuclear push for AI power
AI data centers depend on more than advanced compute. Without sufficient power delivery and generation capacity, new capacity cannot be brought online reliably. The article highlights that electricity availability can be the limiting factor even when hardware capacity exists, noting that a typical 100-megawatt data center can use as much power as about 80,000 U.S. homes.
Against that backdrop, Oklo and X-Energy are effectively competing on a core question investors care about: which company can supply usable nuclear electricity in the time window when data center demand ramps quickly.
Oklo’s Aurora-INL: focus on getting online by 2028
Oklo has moved further through the pre-construction phase, according to the article. It was selected for the Department of Energy’s Reactor Pilot Program, which streamlines aspects of testing and licensing for a designated reactor. That reactor is described as the Aurora-INL powerhouse located in Idaho Falls, Idaho, and the company broke ground in September 2025.
On timing, Oklo’s finance leadership pointed to opportunities to accelerate work tied to the project’s target schedule. On its second-quarter earnings call, CFO Richard Bealmear said the company is “gaining greater visibility” into opportunities to bring forward critical work that supports Aurora INL’s planned 2028 start-up, according to the article.
The target 2028 start-up is presented as a central thesis because, as the article notes, an estimated 2,441 data centers are expected to be online by December 2028. For investors, the implication is that early electricity availability could translate into real customer pull and operational data that strengthens the next phase of commercialization.
X-Energy’s licensing approach: slower now, built for scale later
X-Energy’s deployment timeline, as characterized in the article, is less advanced operationally than Oklo’s. It notes there has not been groundbreaking for the project, with no construction activity or site readiness reported in the same terms as Oklo’s.
However, the article draws a key distinction in licensing strategy. Unlike Oklo, X-Energy is not described as being under the Reactor Pilot Program. Instead, it is moving through the Nuclear Regulatory Commission’s standard licensing pathway.
The article frames that route as slower, but it also argues it may be structurally better suited to achieving a full commercial operating license—an important step if the business model requires multiple reactors rather than a single early deployment.
Speed versus scale: how investors may weigh the outcomes
The article ultimately treats the competition as a trade-off between speed and scale. Oklo’s approach is presented as potentially better positioned to reach power generation first at Aurora-INL, which could matter disproportionately if thousands of data centers need electricity on tight timelines by the late decade.
At the same time, the article notes that even if Oklo begins operations in 2028, future deployments would still require additional regulatory approvals for each new deployment. That keeps execution risk on the table for both companies, regardless of which one starts producing electricity first.
For X-Energy, the central bet is different: while approvals may take longer, the standard NRC licensing path could support a more straightforward path to building a larger fleet once the regulatory process is complete. In practice, that could position X-Energy more favorably for customers that plan for long-term power procurement across many facilities.
What to watch next
Investors tracking the AI power buildout may want to monitor project milestones and regulatory progress that confirm each company’s stated timeline. For Oklo, additional updates tied to Aurora-INL’s 2028 start-up plans will be key; for X-Energy, progress through standard licensing steps and any movement toward construction readiness will likely determine whether the company can convert approvals into a broader reactor program. Upcoming earnings updates and regulatory developments will be the next catalysts to watch.
