There is a wide gap between AI’s hype and its use in business

America’s biggest companies can’t stop hyping artificial intelligence to investors, but new data shows few businesses are putting the technology to actual use.

Nearly half of S&P 500 companies have talked about AI during earnings calls since May, an NBC News analysis of S&P 500 earnings calls found. In fact, AI was mentioned about as often as the Federal Reserve and interest rates. But the hype isn’t translating into real-world use: A November survey from the Census Bureau found that just 4.4% of businesses nationwide reported using AI to produce goods or services recently.

Interest in artificial intelligence went into overdrive with OpenAI’s introduction of ChatGPT in November 2022. NBC News’ analysis found that mentions of AI during earnings calls skyrocketed in the quarters immediately after the chatbot’s debut.

Companies that have begun lauding AI include Salesforce, whose CEO, Marc Benioff, told investors during its second-quarter earnings call in August that the company was leading its customers into “the new AI era.” Salesforce’s investment arm also launched a $500 million generative AI fund to invest in businesses that develop AI models, a list that includes Anthropic, Cohere and Humane.

Businesses outside of the tech world are getting in on the trend, too. In November, Walmart told investors that it is now testing AI-powered search and chat features, and Bath & Body Works announced the pilot of a machine-learning tool that the company hopes will help it win back inactive customers.

“There’s not a single call I sit on these days where AI is not mentioned in some form,” said Brian Nagel, a senior analyst at the investment firm Oppenheimer & Co. who specializes in consumer and e-commerce companies. But from his vantage point, Nagel added, the actual implementation of AI by businesses is “conservative.”

In fact, only a small fraction of these industries have actually put AI to use in creating the products they sell, according to the Census Bureau survey of 200,000 businesses.

Why the divide? Nagel said that, to some extent, companies publicly lean into hype to show investors that they are committed to long-term growth. At the same time, the technology is in its “very, very early stages,” meaning that businesses used to working one way for a long time can struggle to adapt or find practical use cases.

A deeper explanation may be resources — or, rather, a lack thereof.

“When you get some exciting, new technological change, it does not magically appear in the economy,” said Kristina McElheran, an assistant professor at the University Toronto who has studied how technology affects business strategy and performance. Instead, McElheran believes that it is education, access to talented employees, and money that help companies make use of new technologies.

McElheran released a study in October on AI adoption in business that found few businesses are able to use AI tools currently, as the tools are expensive and relatively few people know how to work with them.

“It’s a double-edged sword. The things that can really promote progress and prosperity can also leave certain groups behind,” McElheran said.

Currently, McElheran said, the imbalance is reflected in early adopters clustering around what she terms “superstar” cities — which she says highlights a coming “AI divide” between businesses and cities that can use AI tools and those that can’t.

Mike Loukas, CEO of TrueMark Investments, is more bullish on the wide use of AI in work, predicting that “north of 75%” of businesses will use AI in the next five to 10 years. The caveat is that few of those businesses will work with AI directly. Instead, Loukas believes, they’ll use applications built on top of AI that look familiar and accessible.

For instance, a doctor’s office with a handful of employees might not develop an algorithm to personalize its patient portal, but it might use a questionnaire tool that incorporates artificial intelligence to do the job.

“AI is not going to be doing the work for” the average business owner, Loukas said. “But there will be applications that allow [them] to run his business more efficiently and smoothly.”

CORRECTION: (Dec. 4, 2023: 10:52 a.m. ET): A previous version of this article misstated Kristina McElheran’s position at the University of Toronto. She is an assistant professor, not an associate professor.