U.S. tech firms have been deepening ties to the Middle East with Gulf states investing heavily in AI.

But as this geopolitical turmoil in the region heats up, here DeBos is looking at how those ties are starting to expose the risks of that so-called sovereign AI bucket.

We talked so much about last month, Dee.

So much about it.

Remember, Carl, that sovereign AI, it was supposed to be the next big thing,

a wave of national efforts across the Middle East and beyond, picking up where China left off

after U.S. exports control kicked in. Sovereign AI is countries that are trying to build their

own models on their own data inside their own borders. But the money hasn't always come

with a clear business case,

especially in the Middle East. In many cases, they were tied to government-led projects,

built for political prestige rather than commercial demand. And that is the risk here. Unlike

hyperscalers or global enterprises, government priorities, they can shift fast or they can even

pull funding if geopolitical

tensions escalate. For investors, this could mean a repricing across the AI trade, one that

has been powered by relentless demand and reliable buyers. Sovereign AI spending in the Gulf,

it breaks that model. It's potentially massive, yes, but it's not reliable. And if that

spending stalls or if geopolitical

risks escalate, the sovereign AI tailwind, well, it could turn into a headwind for the names

that have benefited the most, like Nvidia, Microsoft, Open AI, Amazon, others. One area, though,

where spending looks more durable is defense. Palantir, of course, it's up nearly 90% this year. And then you got

Andrel. Earlier this month, it's still private company, so it raised new capital at a more

than $30 billion valuation. These two have been riding a different kind of sovereign wave,

...