Cramer wants proof that AI spending is boosting profits

Jim Cramer said Wednesday that companies need to start proving artificial intelligence is improving their financial results. For everyday investors, the issue is whether the AI boom is producing real savings and revenue growth for customers, or mainly lifting the companies that sell the hardware and infrastructure behind it.

The CNBC “Mad Money” host said he still sees a long-term opportunity in AI, but wants clearer evidence during earnings season. “I need cold hard return facts,” Cramer said. “Or, I, too, will grow more skeptical than I am now.”

AI has driven a sharp rise in spending by major technology companies. Analysts cited by CNBC estimate total AI capital expenditures could top $1 trillion in 2027. Capital expenditures, or capex, are long-term investments in assets such as data centers, chips and networking equipment.

Cramer’s concern is that many companies using AI have not yet tied the technology to higher sales, lower costs or improved profit measures. He said early earnings reports have not included much concrete evidence that AI adoption is changing company numbers in a meaningful way.

Infrastructure sellers are seeing the benefit

Cramer drew a line between the companies building AI systems and the businesses buying them. He said AI infrastructure providers and component makers continue to benefit from heavy spending, pointing to companies such as Anthropic and memory-chip maker Micron, whose profits he said have climbed.

The less clear part, according to Cramer, is the payoff for end customers. He questioned why more corporate users cannot identify at least modest savings from AI projects if the technology is already making operations more efficient.

Large cloud-computing providers, often called hyperscalers, are central to the AI buildout because they operate the data centers that run AI workloads. Cramer said investor confidence in those companies could weaken if customers do not start showing how AI helps them make or save money.

Banks have not shown enough AI impact, Cramer says

Cramer singled out banks as one area where he expected more progress. He said financial institutions seemed well positioned to use AI for automation and efficiency, but their management teams have not provided enough evidence that the technology is materially improving results.

He said AI may still be useful inside banks, but added that he has not seen proof that it is improving efficiency ratios or reducing hiring needs. An efficiency ratio is a common bank metric that compares expenses with revenue, with a lower ratio generally signaling a more efficient operation.

“Does that mean AI is a bust? No,” Cramer said. “But I don’t see it making much difference.”

Cramer said only a small group of companies, including fintech firm Block and web-security company Cloudflare, have clearly connected layoffs to AI adoption. CNBC reported that Block did so in February and Cloudflare disclosed job cuts in May.

He also noted that critics say some companies may use AI as a trendy explanation for layoffs, a practice sometimes called “AI washing.” In that case, companies may imply AI is driving a business change without showing that the technology is actually responsible.

Cramer said the burden is now on corporate customers to show measurable financial gains. If they do not, he said, skepticism around AI spending is likely to build, especially around the biggest technology companies funding the buildout.

This story draws on original reporting from CNBC.