Most CEOs aren’t buying the hype on generative AI benefits

Generative AI concept art featuring a glass human brain on digital background
(Image credit: Getty Images)

Most CEOs think it will take at least two years for generative AI to deliver tangible business benefits, with most executives happy to make small-scale experiments with the technology for now.

A survey from the Boston Consulting Group found that two-thirds of the executives it spoke to believed it will take at least two years for generative AI to move beyond hype, and 71% are focused on pursuing limited experimentation and small-scale pilots. In total about 90% of executives were in this category.

They are not alone in this, either. In August last year, tech analyst Gartner placed generative AI at the ‘peak of inflated expectations’ on its annual tech hype cycle graph, highlighting that while there is a lot of excitement around the technology, it could take two to five years before it hits the mainstream.

Despite worries about hype, the survey did find that companies are ramping up their AI spending, with 71% of executives planning to increase their company’s tech investments this year and 85% intending to increase their spending on AI specifically.

Alongside this, 89% of the 1,400 execs surveyed ranked AI as a top-three tech priority for 2024 alongside cloud computing and cyber security.

Generative AI progress causing friction

Many CEOs told BCG they were dissatisfied with their organization’s progress on AI, blaming a lack of talent and skills, as well as an unclear roadmap and investment priorities. 

Only 6% of companies have trained more than 25% of their staff on generative AI tools so far, and nearly half (45%) of executives said that they don’t yet have guidance or restrictions in place to regulate the use of generative AI in the workplace.

54% of leaders already expect AI to provide cost savings this year, primarily through productivity gains in operations, customer service, and IT.

But while most companies are holding back, those that are spending big on generative AI are perhaps unsurprisingly the most likely to anticipate benefits.

Organizations that plan to invest more than $50 million in AI are 1.3 times as likely as their peers to expect cost savings in 2024 — and 1.5 times as likely to anticipate more than 10% in cost savings, the survey found.

Execs optimistic about business benefits

A separate report from PwC paints a slightly more optimistic picture about the current outlook on generative AI. 

Around one-third of the 4,702 CEOs it surveyed said that generative AI had been adopted across their organization in the last 12 months, or that their organization had changed its IT strategy as a result of AI.

Over half expected AI to improve the quality of their company’s products or services over the next year, and two-thirds said generative AI would require most of their workforce to learn new skills in the next three years.

That’s assuming those workers still have jobs anymore, however. One-quarter of CEOs said they expect to reduce headcount by at least 5% in 2024 due to generative AI, although PwC pointed out that companies making cuts in some areas may already be offsetting them with hiring in others.


A webinar from IBM to help you take your hybrid cloud journey to the next level with generative AI

(Image credit: IBM)

Take your hybrid cloud journey to the next level with generative AI


Media, banking, insurance and transport CEOs were among those most likely to say that generative AI will lead to job losses in their sector.

“Ultimately, CEOs must embrace this as a new facet of their role: understanding, explaining and managing the inevitable tensions between short-term job losses and long-term job creation potential from AI,” said PwC.

PwC noted that leading companies are also aligning their generative AI adoption efforts with existing digital and AI strategies, upskilling employees, and encouraging experimentation across their organizations.

“In these early days of generative AI adoption, most companies are still nailing down what they’re trying to accomplish — and why — with this powerful, general purpose technology,” the report stated.

“As well they should. Despite the enthusiasm, generative AI is only one type of AI, and has yet to achieve anything close to its potential.”

Steve Ranger

Steve Ranger is an award-winning reporter and editor who writes about technology and business. Previously he was the editorial director at ZDNET and the editor of