If 2023 was the year of generative AI hype, 2024 will be the year during which most corporations will get serious about spending on the technology to bring artificial intelligence more directly into the lives of their workers and customers.
More than half (55%) of top technology officers across sectors of the economy say they plan to purchase enterprise-level gen AI software, such as Microsoft Copilot, in the next six months. While another third (32%) said they have not made the spending decision, only 13% said they would not be acquiring similar gen AI capabilities.
That's according to the CNBC Technology Executive Council Survey for the second half of 2023, conducted in late November and early December among a sampling of 22 top tech officers. More than half (59%) of the respondents, which include chief technology officers, chief information officers and chief cybersecurity officers, say their company's new investments in artificial intelligence capabilities are accelerating. The remainder of the executives are evaluating new investments in AI, but are being more cautious. No respondent said they will not be making new AI investments in 2024.
For Microsoft, which has assumed an early lead in the generative AI market as rivals like Alphabet play catchup with its Google Bard, the spending plans should be good news. It's clear Wall Street has already baked in some big gains for the company, whose shares are up 55% this year, more than doubling the S&P 500 return year-to-date.
"This is a clear positive as right now Microsoft has the only game in town around a full AI platform," said Wedbush Securities analyst Dan Ives. "Enterprises are heading aggressively down this path and Copilot is the torch bearer for the broader AI market."
Analysts have put price targets as high as $600 on Microsoft over the the next three years and predict gen AI revenue to reach as high as $10 billion annually. OpenAI has received billions in investment from Microsoft and since ChatGPT's introduction in late November 2022, Microsoft shares have increased from roughly $240 to over $370. That's despite the recent boardroom drama which threatened the startup's existence — Microsoft is now placing itself more directly in a position to monitor the startup's operations.
"With all the hype around generative AI, coupled with Microsoft's heavy marketing push on Microsoft 365 Copilot, it is not surprising that it tops the list," said Jason Wong, a Gartner analyst.
He cited a recent Gartner survey that found 82% of IT buyers saying Microsoft 365 Copilot was in their top three among new 365 features that will be "most valuable" to their organizations.
Corporate spending is due to accelerate because of Microsoft's general release of Copilot for 365 in November. An early pilot program which started last May and reached 600 customers, according to Microsoft, was passed over by many Gartner clients, Wong said, due to concerns, primarily about cost, but also potentially due to data and legal worries.
Some corporate buyers are still balking at the requirement of a 300-seat minimum for the software that costs $30 per user per month, Wong said, but now that Copilot is generally available, many companies that skipped the early pilot are making their move to adopt the gen AI. In August, Alphabet launched Duet AI for Google work apps at $30 per month for corporate users.
Microsoft has pointed to customers including Visa, BP, Honda and Pfizer using Copilot, and professional services firm partners on Copilot AI including Accenture, EY, KPMG, and PwC. Some firms are already deploying the AI on a mass scale, including PwC, for example, which is rolling out ChatPwC to 75,000 workers by the end of this year.
"The hyperbole around AI is more than I have seen with anything in 30 years, but it's well-earned. It will change everything," PwC chief products & technology officer Joe Atkinson, a member of the CNBC Technology Executive Council, recently told CNBC.
At that level of deployment, the $30 per user price tag would equate to tens of millions of dollars in annual spend. But Gartner's research suggests that within big companies, the initial spend won't necessarily reach as far across the workforce. "Anecdotally, through the thousands of inquiries we get on this, companies are very rarely willing to just go wall-to-wall from the get-go," Wong said.
How companies will begin to spend billions on AI
The expectation, according to Gartner, is for test groups to be the focus at the start over a period of three to six months, such as within IT or specific business functions like sales and marketing, areas where need for AI to solve a core problem already has been identified. "That will be the first half of 2024 for the majority of buyers," Wong said.
Ultimately, Gartner anticipates that the 300-seat minimum will be eliminated, but Wong said "Microsoft wants organizations to put their skin in the game," and in fact, that is the way to get the most out of AI and will push companies to quickly extend the availability across the workers. "The more content and knowledge that is in across users, the better AI will perform," he said. "If it's just ten users it won't be very helpful, so organizations need to take into account with limited seats what they can really do with it," he said.
Productivity savings will be identified even in smaller groups. That's been the case to date. Microsoft's recent Work Trend Index report was the first to look at Copilot early usage and found that 77% of people who have used Copilot said they don't want to give it up; 70% of Copilot users said they were more productive, and 68% said it improved the quality of their work.
It's not just Microsoft blowing smoke. A CNBC survey of thousands of employees across the U.S. workforce released this week found that 72% of workers using AI say it has made them more productive.
Nearly half of the tech executives surveyed by CNBC say that up to 25% of their employees already are using gen AI in their work. The CNBC TEC survey also found that companies are balancing their AI spend between the worker and customer. Roughly one-third indicated more focus on AI for the customer, while one-quarter said the primary focus is on AI for the workforce.
If this is the start, Gartner says the end of the road for enterprise deployment is broad connectedness to other data sources outside 365 work apps specifically, including ERP, CRM, supply chain, financial and legal systems, "all the other places where knowledge is," Wong said. "It's not just generating emails and creative Powerpoints. To get real value, you have to integrate other sources into Copilot and that's where the enterprise-wide deployment will really accelerate," he said, though he added it's still in the early days of making that a smooth process for companies, with all of the apps and plugins available and up to the task and compliance-ready.
All the major enterprise software vendors, from Salesforce to SAP, Workday, and Adobe will be rolling out products with embedded gen AI features and will also be charging more for the features, so organizations will quickly have to decide what these offerings do for employee experience, and what features will drive adoption. Even at Microsoft, the tech giant is making sure to embed third-party AI on its Azure cloud platform, such as Mistral, as its chief rivals including Amazon Web Services work with multiple AI vendors. Microsoft is also developing its own open-source AI, Orca, which has been the primary to-market approach for Meta.
All the offerings could result in an initial overspend, or overly cautious budget from others, in a market that within the U.S., at least, includes a healthy dose of FOMO and organizations already asking "Will I fall behind?"
And even amid the hype, it's important to keep in mind that as gen AI spending grows, it is still dwarfed by companies' cybersecurity budget needs. In 2024, Gartner predicts that $114.8 billion dollars will be spent by enterprises on security, versus $21.8 billion dollars on generative AI. That means roughly $5 will be spent on security for every dollar spent on gen AI. For every dollar spent on IT as whole, only one-half of a cent will be allocated to gen AI, according to Gartner's chief forecaster John Lovelock.
But it's growing, and for now at least, when it comes to the billions in gen AI spending, Microsoft is in the pole position. "They have the install base and more important, the amount of data in the Office suite and email and Teams, and SharePoint and OneDrive, that's where they [companies] can really hit the ground running in using gen AI," Wong said, "as long as there is the governance and guardrails."
