What is the rationale for the Tumas FID deferral?

The decision not to proceed with the Final Investment Decision (FID) was not due to any project weakness, but based purely on market conditions, particularly the mismatch between supply and demand: Downstream demand for uranium is completely transformed, yet when we look at supply, nothing has changed. Worse yet, the projects coming online are seriously in trouble, with 5–6 million pounds/year reductions in their combined guidance. So what I am saying—and what I said to our board—is that I am not getting out of bed in a situation whereby utilities are trying to recreate the past; the inventories and uranium prices of “normal” years ago are a dream gone. Utilities, especially US utilities, need to realize that the uranium industry can only be recapitalized and exploration can only be incentivized when uranium prices rise. So my approach, an approach vindicated by others in the industry, is that we will not talk seriously to utilities until the incentive price is enough to justify development of greenfield projects. Market leaders like Cameco have also said no to any greenfield projects without three digits in the spot price (over US$100/lb). The uranium market is currently divided into a spot market for immediate transactional purposes, which accounts for only about 10%-20% of total production, and the term market, which feeds the nuclear fleet with long-term contracts. That term market has been dead, sitting at around US$80/lb, and we need it to be active again; that will be the trigger point for us.

Could you elaborate on the progress at Tumas?

Deep Yellow is well cashed-up, with about US$220 million in the bank, so we were able to continue works without being caught in the tangle of financing. The project has now progressed to Phase Three of detailed engineering, with Ausenco handling the processing plant, while we manage offsite infrastructure, including power, water and roads. Most companies reach FID at 30% engineering completed, primarily on the front end, whereas we’ll be at up to 70% when Phase 3 finishes. We are also implementing innovative construction methods like off-site fabrication and pre-cast concrete to save time and reduce on-site risks. This strategy ensures the project can start construction quickly and efficiently once market conditions justify it.

How do you see the uranium supply gap evolving over the next decade?

Supply sector stagnation since 2012 and weak uranium prices have created a major supply deficit going forward. Current consumption is 180 million lb/y—growing to around 300 million lb/y by 2040—which means we need 130 million lb/y of new resources by the mid-2030s. With 50 million lb/y from current production lost to depleting mines, we must find 200 million lb/y in 15 years. Most greenfield projects are years away from production.

I take a visionary look to understand where nuclear is going and the type of companies the space needs in the 2030s. We have two greenfield projects scheduled to go into production before the end of 2030 (in 2027 and 2029). With our projects, we are looking at a 35- to 40-year life of mine. The reason for that has always been to establish long-term customers, create a stable revenue runway, and provide a foundation for the people who will run Deep Yellow after I’m gone. This will allow them to grow production from 5–7 million lb/y to 15–20 million lb/y, which is where I believe Deep Yellow needs to be.

In what ways are new drivers, like data centers and AI, shaping the uranium demand curve?

Current demand for nuclear is unprecedented. Previously, China led nuclear expansion. Now, after 15 years of stagnant fleet growth, every new reactor being built requires four times the typical annual uranium for its initial core. Vendors must supply a year’s worth of uranium during construction—1,000-megawatt reactors need 1.8–2 million lb initially, then 500,000 lb/y. This pushes demand for both new and existing reactors.

This is without including the hyper-scalers with their data centers, and AI aspirations, which most analysts haven’t yet fully accounted for in their models. The US will require the equivalent of Japan’s entire electricity consumption just to support AI and data centers by 2030. Meanwhile, the scale of the problem is so huge that data center operators will burn anything—coal, gas, you name it—because the competition for economic supremacy is enormous. Eventually, with focus on low emissions, nuclear will begin to dominate.

Do you have a final message?

The problem with the energy transition, or energy crisis, is the oversimplification of the challenge by ideologically driven politicians who think a 300-year Industrial Revolution, with all the capitalization this required, can be transformed in 20 years. Hydrogen and renewables have challenges that are not fully appreciated. 

Transitioning from gas and coal will take longer than many expect. Nuclear, however, remains central, especially for hyper-scalers. Modern reactors are like the pyramids of our days, with lifespans lasting for a century. Nuclear plants don’t become redundant just because a more efficient technology comes along. For example, a reactor built in 1954 was still operating until five years ago.

*John Borshoff stepped down from his role as Managing Director and CEO on October 20, 2025, and will remain an advisor to the company until the end of November.