This is The Takeaway from today's Morning Brief, which you can sign up to receive in your inbox every morning along with:

• The chart of the day

• What we're watching

• What we're reading

• Economic data releases and earnings

Everyone is so gaga for AI these days that whenever I encounter skepticism, it piques my interest.

That’s why Guggenheim’s latest note on IT spending caught my eye. Lead analyst John DiFucci calls for a “new normal” for software companies — a reality in which enterprise budget growth is not going to reaccelerate anytime soon. He based that outlook, in part, on data from tech researcher ETR that shows July IT budget growth expectations are about the same as last year’s.

What about the AI spending boom? What about Nvidia’s parabolic sales increases as companies build up their data center firepower to build their large language models and prepare for the AI age?

That hasn’t made its way to software, DiFucci argues. That’s partly because of cost, and it seems partly because companies haven’t yet figured out what AI is useful for. (As I wrote recently, that’s the case with consumers as well).

Corporations may end up spending on GenAI eventually, just not yet.

“Most of the spending on AI is being done by AI companies and Public Cloud companies preparing to run AI workloads for AI companies,” DiFucci wrote. “That doesn’t mean that we won’t see that shift at some point, when corporations begin to purchase Copilots and other forms of AI en masse, or start to build their own LLMs as the cost to build and train them continues to decline. But that doesn’t seem to be the case right now.”

What’s more, he said, companies don’t seem to be holding off on IT spending today in order to spend on AI tomorrow: “Given the uncertainty regarding what AI spending will be useful to corporations, we do not think that expectations for future AI spending is driving what appears to be a more challenging IT spending environment today.”

The “build it and they will come” approach on the part of companies like Microsoft, Alphabet, and Meta may well pay off down the line.

But despite some analysts’ insistence that this is the year that AI proves itself in practice (Forbes called this “the year of AI practicality”), it’s still unclear exactly how.

To DiFucci, that means about half of the software companies he covers may have to reset lofty AI-related revenue expectations. That includes firms like Palo Alto Networks, Salesforce, and Workday.

At the same time, there are many corporate leaders who will tell you (a little too eagerly, perhaps) how they — and their clients — are already using AI. But as Jefferies’ Brent Thill said earlier this year, it won’t be until 2025 that that revenue will start to materialize.

If corporate number-crunchers are keeping the purse strings tight, AI may have to prove its worth before they start spending.

Julie Hyman is the co-anchor of Yahoo Finance Live, weekdays 9 a.m.-11 a.m. ET. Follow her on Twitter @juleshyman, and read her other stories.

Click here for the latest technology news that will impact the stock market

Read the latest financial and business news from Yahoo Finance