It's a gut-wrenching feeling when a stock you're bullish on suddenly tanks. It's downright terrifying when the forces that were supposed to lift that stock higher all seem to vanish at once.
With that in mind, let's run a scenario on Oklo (OKLO -5.50%). Could this early-stage developer of advanced nuclear reactors, which has tanked about 80% since last October, survive without an AI boom?
NYSE: OKLO
Key Data Points
What could happen to Oklo if AI demand disappoints?
On the one hand, Oklo could survive without an AI boom. On the other hand, its bull case would not.
A quick glance at Oklo's project pipeline would suffice to illustrate what I mean. Its 18-gigawatt (GW) backlog is anchored by two major agreements: Switch's, up to 12 GW, and Meta's, up to 1.2 GW. If we throw in Equinix's at 500 megawatts (MW), then about 93% of Oklo's backlog is associated with AI data centers.
Image source: Oklo.
None of these are binding agreements. An AI bust, by implication, could easily leave this non-diversified pipeline looking like a creek-bed in a dry spell.
In such a scenario, revenue growth from non-AI-related demand would be much harder to scale. Think about it like this. A single hyperscaler can absorb hundreds or thousands of megawatts (MW) of electricity, which requires a dozen or so of Oklo's 75-MW powerhouses to match it. If multiple data centers belong to the same operator, such as Equinix or Switch, then Oklo could potentially secure hundreds of millions or even billions of dollars in annual revenue through one relationship. As you can imagine, a few of these commercial partnerships could open the throttle on Oklo's growth.
If all our AI hopes and dreams flop, Oklo's fast-fission reactors could still attract clients, like chemical factories, military camps, and utilities. But none of these would likely need more than one or a few of Oklo's 75-MW powerhouses. The opportunity would be there, but it wouldn't be as big as the needs of a data center.
A bright spot, and a caveat
That said, one non-AI bright spot for Oklo right now is its isotope business.
These special materials are used in cancer treatment and diagnostic imaging, among other things, and global supply is very constrained. Oklo's subsidiary, Atomic Alchemy, was recently granted a Nuclear Regulatory Commission (NRC) license for isotope material, which basically means it can start selling recovered and prepared materials from its Idaho laboratory. The larger point: Oklo could soon have a source of revenue that's not tied to AI, one that might even arrive before its first powerhouse begins generating electricity.
Still, even with a thriving radioisotope business, Oklo needs the opportunity from AI to justify its current valuation. It carries a market cap above $7 billion, which is mid-cap territory, yet it generated zero revenue in 2025.
When measured against Wall Street's revenue estimates, the valuation looks even more absurd. Two fiscal years from now, projected revenue is about $55 million, which means Oklo stock trades at about 127 times forward sales. That leaves absolutely no room for an AI bust, not even a small one.
Data by YCharts.
To get back to the question at hand: No, Oklo would not be an attractive buy if AI turns out to be a bust.
Without hyperscale customers, Oklo's order book would evaporate, and it would need to have binding agreements with industrial and defense clients to restore even a modicum of confidence. Likewise, radioisotopes are an exciting side venture, but if they turn into Oklo's main sale, the business's total addressable market (TAM) will shrink meaningfully.
Like other energy stocks connected to AI, Oklo needs the technology to justify its valuation. That tenuous relationship is why this stock will tank on any negative AI news -- and why investors with a weak stomach for volatility should probably stay away from it for now.
