Intel Just Got a Rare Double Upgrade From Bank of America. Here's the AI Shift Behind the Call.

Most analyst rating changes move a single notch. So when one of the more closely followed semiconductor analysts skips the middle rating and jumps two rungs at once, it's worth a closer look -- not necessarily as a reason to buy, but as a window into how the thinking around artificial intelligence (AI) spending may be changing.

That's what happened on June 11, when Bank of America's Vivek Arya double-upgraded Intel (NASDAQ: INTC) to buy from underperform, lifting his price target to $135 from $96. Banks tend to reserve a move like that for moments when the thesis they had been betting against breaks. And it came after Intel had already more than tripled in 2026 -- less a call on a forgotten stock than a bet that the turnaround has further to run.

Intel shares rose about 6% in the following session, closing near $125 as of this writing.

The more useful question isn't whether the target is right. It's what the upgrade says about the next phase of the AI trade.

Image source: Getty Images.

The case for the CPU

For most of the AI boom, the story has been about graphics processing units (GPUs) -- the chips that train large models. Central processors, Intel's historical stronghold, were an afterthought. The upgrade rests on the idea that this is starting to change.

The argument goes like this: as companies build "agentic" systems in which software agents carry out tasks independently, incremental work shifts to the central processors to coordinate those agents.

If that shift plays out, the market for server processors will get much bigger. Indeed, Bank of America now pegs the market at more than $170 billion by 2030, up from a prior estimate of around $125 billion, and sees Intel capturing roughly a quarter of it.

Intel's own recent numbers lend the idea some support.

In the first quarter of 2026 (the period ended March 28, 2026), the company's data center and AI revenue rose 22% year over year to about $5.1 billion -- easily the fastest-growing of its three main segments, well ahead of the 1% growth in its client computing group, which sells PC chips.

And management is describing the central processor in much the same terms.

"In recent months, we have seen clear signs that the CPU is reasserting itself as the indispensable foundation of the AI era," said Intel CEO Lip-Bu Tan in the company's first-quarter earnings call.

A risky foundry business

The second part of the upgrade is about manufacturing.

Intel is trying to become a more formidable contract chipmaker, producing chips for others the way Taiwan Semiconductor does -- a business it calls its foundry. This, however, is the higher-risk, higher-reward part of the story. Bank of America pointed to a pipeline of potential outside customers, including reported work on Apple's chips and MediaTek's processors. If even a few of those deals scale, the foundry could grow from a rounding error into a real second engine.

But Intel isn't there yet. Intel's external foundry revenue -- the part that comes from outside customers rather than its own products -- was just $174 million in the first quarter. And the foundry segment as a whole is still losing billions of dollars a year as Intel spends heavily to build out advanced manufacturing. And closing the gap with Taiwan Semiconductor is something Intel has tried and stumbled at before.

Then there's the price. Intel trades at well over 100 times its expected earnings for the next year. A valuation like this leaves little room for error, assuming both the agentic-processor shift and the foundry turnaround unfold with very few blips along the way.

So, what do I make of all this?

The rotation Bank of America is describing looks real and underappreciated -- the idea that AI's next phase pulls spending back toward the processors and manufacturing capacity Intel is built around. But a compelling thesis and an attractive entry price are different things. After more than tripling this year, the stock has arguably already priced in much of that optimism, and the bar it now has to clear is far higher than it was a few months ago. I'd want to see Intel's foundry business make more meaningful progress on its pipeline before treating the stock as a buy here. Sure, the shift may be coming. But the stock apparently isn't waiting around for proof.

Bank of America is an advertising partner of Motley Fool Money. Daniel Sparks and his clients have positions in Apple. The Motley Fool has positions in and recommends Apple, Intel, and Taiwan Semiconductor Manufacturing. The Motley Fool has a disclosure policy.