Cramer presses companies for evidence of returns on AI spending

CNBC’s Jim Cramer said companies using artificial intelligence need to show measurable financial returns as spending on the technology continues to climb, with analysts cited by CNBC estimating AI capital expenditures could top $1 trillion in 2027. He said infrastructure providers and component suppliers are benefiting, while many corporate customers have yet to report material savings, revenue gains or profit improvement from AI adoption.

Speaking Wednesday on CNBC’s “Mad Money,” Cramer said he remains constructive on the longer-term opportunity in AI but wants more evidence in earnings reports that the technology is affecting corporate results. “I need cold hard return facts,” he said. “Or, I, too, will grow more skeptical than I am now.”

The distinction matters for investors because the economics of the AI buildout can show up unevenly across the supply chain. Companies building and supplying AI infrastructure can book sales as large cloud and technology groups spend on chips, memory, data centers and related systems. The customers deploying AI inside their own operations must then prove that those tools either cut expenses, raise output, improve revenue or support margins.

Cramer said that proof has been limited so far in the current earnings season. “We’re still early in the earnings season but already we are not hearing anything material about the use of AI,” he said.

Banks draw scrutiny over efficiency claims

Cramer singled out banks as an area where he expected stronger evidence of AI-related benefits. He said financial institutions appeared to be logical adopters because AI could automate workflows and improve efficiency, but he said management teams have not yet shown that the technology is changing reported performance in a meaningful way.

He characterized the benefits described so far as useful but insufficient to alter financial expectations. “It’s valuable, but nothing that can raise numbers,” Cramer said. “It’s not helping the efficiency ratio that we can tell and it’s not allowing them to cut back on hiring. Does that mean AI is a bust? No. But I don’t see it making much difference.”

The efficiency ratio is a closely watched measure for banks, comparing operating expenses with revenue. If AI reduces staffing needs or processing costs, the effect would be expected to appear through that kind of metric, although Cramer said he has not yet seen clear evidence in reported results.

Suppliers benefit before customers prove gains

Cramer said AI infrastructure companies and some suppliers continue to see benefits from the spending cycle. He pointed to Anthropic as a company getting returns and cited component companies including Micron, the memory-chip maker, as beneficiaries of strong demand.

His concern is centered on the end users paying for AI tools and infrastructure. “But shouldn’t the ultimate clients ... be able to cite at least a couple of million in savings?” he said.

Cramer said only a limited group of companies, notably fintech firm Block and web-security provider Cloudflare, have clearly linked job reductions to AI adoption. CNBC reported that Block disclosed cuts in February and Cloudflare disclosed cuts in May. Cramer also noted that critics say some companies may invoke AI as a convenient explanation for layoffs, a practice CNBC said has contributed to use of the term “AI washing.”

Without more evidence from customers, Cramer said skepticism about the AI trade is likely to intensify and could affect how investors assess spending by large technology companies. “The longer we go without hearing how actual clients make money, the longer we’ll take days like today, when it seems that the hyperscalers are making money,” he said, “with a grain of salt.”

This story draws on original reporting from CNBC.