- Generative AI is unlike any technology that has come before. It’s swiftly disrupting business and society, forcing leaders to rethink their assumptions, plans, and strategies in real time.
- To help CEOs stay on top of the fast-shifting changes, the IBM Institute for Business Value (IBM IBV) is releasing a series of targeted, research-backed guides to generative AI, on topics from data cybersecurity to tech investment strategy to customer experience.
- This is part six: Tech spend.
Generative AI is unlike any technology that has come before. It’s swiftly disrupting business and society, forcing leaders to rethink their assumptions, plans, and strategies in real time.
To help CEOs stay on top of the fast-shifting changes, the IBM Institute for Business Value (IBM IBV) is releasing a series of targeted, research-backed guides to generative AI, on topics from data cybersecurity to tech investment strategy to customer experience.
This is part six: Tech spend.
Generative AI is unlike any technology that has come before. It’s swiftly disrupting business and society, forcing leaders to rethink their assumptions, plans, and strategies in real time.
To help CEOs stay on top of the fast-shifting changes, the IBM Institute for Business Value (IBM IBV) is releasing a series of targeted, research-backed guides to generative AI, on topics from data cybersecurity to tech investment strategy to customer experience.
This is part six: Tech spend.
Open the aperture to see the full picture of IT spend.
There’s no “one-and-done” generative AI investment. This revolutionary technology promises to impact every business function—and the entire IT estate.
As CEOs rethink business models, job roles, and workflows to capitalize on the full potential of generative AI, they must also carefully consider the broader IT cost implications—and not all of them are obvious.
Leaders need greater visibility into IT spend to better understand and orchestrate the flow of funds across the enterprise. Taking a wide-angle view of data can help them re-evaluate workflows, processes, protocols—and potentially the system architecture itself.
This process may reveal the need for big changes—but will businesses have the IT budget to react? Only if they prioritize projects that will deliver a competitive edge, rather than spreading generative AI spend like peanut butter across the IT portfolio.
The IBM Institute for Business Value has identified three things every leader needs to know:
1. The rush to generative AI is driving unexpected spending.
2. People costs could torpedo your AI strategy.
3. You can’t cut your way to growth.
And three things every leader needs to do right now:
1. Don’t get sidelined by a ballooning IT budget.
2. Unclog the people cost bottleneck.
3. Grab a pencil and the back of an envelope, then connect the dots to build a better business case.
An updated version of this chapter with 2024 data is now available in the latest edition of the IBM Institute for Business Value's book, The CEO's Guide to Generative AI.
Additional content
Download report translations
Originally published 26 September 2023
1. Spending
+ Generative AI
What you need to know
The rush to generative AI is driving unexpected spending
Generative AI is exposing cracks in old-fashioned funding practices. Like any nascent technology, it’s dynamic by nature. What it needs—and how it can deliver value—is changing by the day. And that throws a wrench into the traditional budgeting process.
If leaders can’t be sure which generative AI projects will be most important in the next quarter, let alone the next fiscal year, they won’t be able to allocate funds efficiently. This innovation is poised to disrupt IT budgets—and our data shows the foundation is already shifting.
IT executives now expect 2023 generative AI budgets to be 3.4x greater than they anticipated as recently as four months ago. For a $20 billion organization, this translates to a $5 million increase in projected spend in just over a quarter. And we expect these projections will continue to rise as generative AI matures and use cases crystalize.
However, only 15% of tech execs expect to fund this uptick with net-new spend. Instead, many plan to plunder other parts of the IT portfolio, with 33% saying the money will come at the expense of non-AI IT spend. A slightly larger portion (37%) expect to pull generative AI spending from the broader AI investment portfolio, which could reflect expected synergies across traditional and generative AI projects, some culling of the AI project portfolio, or both.
This approach to reallocation is reasonable, but is it realistic? As generative AI is rolled out across the business, it will have cost implications across the board. Labor and cloud spending, in particular, will grow in lockstep with the increasing demand for generative AI solutions. Overall, the impact is likely to be far-reaching—and a $5 million budget increase won’t get a leading enterprise very far.
What you need to do
Don’t get sidelined by a ballooning IT budget
CEOs need a clear understanding of how high-impact projects will tap resources—both human and technical—to accurately budget for associated costs.
- View spending with a wide-angle lens. Assess the entire web of IT costs required to deliver the impact you expect from generative AI. Analyze spend holistically—across IT, cloud, and people—to deliver enhanced business value across all investments.
- Extend FinOps capabilities across the enterprise to gain visibility into costs and spending across all AI, hybrid cloud, and application modernization investments. Understand what your people are working on and how much they cost, and map that back to specific projects, applications, and initiatives to optimize spend.
- Keep GPUs on your radar. Generative AI requires the extreme processing power of graphic processing unit (GPU) chips, which are in very short supply. The market price for GPUs will drive the cost of building and delivering generative AI services in general and is likely to show up in your enterprise cloud costs.
2. People
+ Generative AI
What you need to know
People costs could torpedo your AI strategy
While generative AI is already more intuitive than many hype-cycle innovations, companies need internal expertise to gain a competitive advantage.
But deep generative AI experience is scarce, which makes talent expensive.
The Wall Street Journal reports that senior AI engineers are fetching up to $900,000 salary offers, with entry-level prompt engineers starting at $130,000. And they’re looking for work that will elevate their resume—not pull them into the weeds.
If companies want to bulk up their bench, they’ll have to be willing to pay—and create positions that offer employees the purpose, autonomy, and opportunities for mastery that they seek. Yet, IT executives are still budgeting for the status quo.
In fact, they’re forecasting declining labor costs overall, with new staffing projected to hit 18% of generative AI spending in 2023, but only 16% in 2025. This may be wishful thinking, as 72% of CEOs have yet to assess the impact of generative AI on the workforce.
As CEOs prioritize specific use cases, the related labor spend calculus is prone to shift. Each generative AI model comes with its own set of labor costs, which means net-new costs will vary for each implementation. This puts leaders in a tough position, forcing them to estimate the financial impact of job roles that don’t yet exist.
What you need to do
Unclog the people cost bottleneck
Bolder, high-ROI initiatives can attract top talent and can help absorb spiking AI talent costs—if your organization can stomach the higher price tag.
- Get an unvarnished view of the current market for talent and the types of talent your organization can compete for successfully. Inner-source to secure skills and flexibility by building a marketplace where those with the right skills (or the interest in learning them) can be matched with opportunities.
- Build market-based people costs explicitly into all generative AI business cases. Model the feasibility of business cases based on the attractiveness of the work, not just on the expected talent costs.
- Engage with strategic partners, especially technology providers and Global Systems Integrators (GSIs), to determine which parties can contribute the specific people required to design and execute your generative AI strategy.
3. Strategy
+ Generative AI
What you need to know
You can’t cut your way to growth
Running a tight ship can cut millions of dollars from a bloated IT budget. But streamlining spend can only take a business so far.
To deliver the exponential returns CEOs dream of, leaders must identify which use cases will drive the most transformative growth. For example, only 2% of executives expect to gain an edge by subscribing to public generative AI services that employees can use, such as ChatGPT, while 38% say using a vendor’s platform with their own proprietary data will deliver that advantage.
Yet, rather than focusing on income-generating areas of the business, organizations are spreading generative AI funding equally across several cost centers. Almost three quarters (74%) of generative AI spending will go to HR, finance, customer service, sales and marketing, and IT, where investments are expected to cut costs. Only 26% will go to product-related business functions, where growth-driving innovations incubate.
This makes it difficult to define business cases that break the mold. To grab the brass ring, CEOs need to make data-driven decisions about which generative AI plays do the most to advance strategic objectives—and fund them accordingly. At the same time, don’t let the perfect get in the way of the good. Proving one or two quick wins can help build the business case for more grandiose visions.
What you need to do
Grab a pencil and the back of an envelope, then connect the dots to build a better business case
Dive into the data to decide where your generative AI program can provide the most bang for your buck. Worry less about financial precision until you’ve designed initiatives worth doing.
- Appropriate the private equity playbook. Apply lessons from the way private equity firms invest in IT. Ruthlessly eliminate initiatives that won’t improve the value of the enterprise within three years—and funnel those funds to programs that will.
- Radically modernize methods and practices for designing strategic IT investments. Allocate spend based on overall growth potential, not just short-term savings. Stop “peanut butter spreading” generative AI funding equally across organizational silos.
- Think ecosystem always. Engage your strategic IT service providers and your most valuable customers in discussions about how to maximize the value of generative AI. Make business model innovation a group project.
The statistics informing the insights on this page are sourced from two proprietary surveys conducted by the IBM Institute for Business Value in collaboration with Oxford Economics, as well as one external reference from the Wall Street Journal. The first survey was answered by 136 US-based Chief Information Officers and Chief Technology Officers in August–September 2023 regarding their perspectives on generative AI’s impact on technology spend. The second survey was answered by a broader range of 369 executives across the US, UK, India, Singapore, Germany, and India in April–May 2023 regarding perspectives on generative AI more generally.
Bookmark this report
Share
An updated version of this chapter with 2024 data is now available in the latest edition of the IBM Institute for Business Value's book, The CEO's Guide to Generative AI.
Additional content
Download report translations
Originally published 26 September 2023
How can IBM help you?
You might also like
You might also like
