Businesses face complex cost-cutting options with GenAI
Some GenAI cost saving candidates might disappoint. The best approach to cost reductions vary according to the business function and an enterprise's strategic business priorities.
When generative AI became mainstream in late 2022, it didn't take long for businesses to explore its cost savings potential.
GenAI's ability to do things faster, such as create content and write code, looked like an obvious efficiency win. Productivity improvement and the potential to reduce costs soon became central to justifying investments in GenAI.
Yet three-plus years after GenAI's enterprise arrival, the results are ambiguous. When looking at operating costs, of which labor tends to be the largest contributor, frequently cited candidates for savings and measurable ROI include business functions such as software development, customer service and marketing. Here, cost savings stem from reducing headcount or cutting back on external service providers.
But actual cost reductions don't always live up to their potential. Some targets fail to deliver the expected savings while others, perhaps unsung, offer better results.
In addition, chief AI officers and other C-suite executives must decide what type of cost reduction will provide the greatest benefit. Operating costs look at an enterprise's ongoing, overarching expenses, with a focus on profit margin. But unit cost -- the cost of producing a single unit of a product -- is another lever available to business decision-makers. Reducing unit cost lets businesses produce more products, accelerate delivery and expand markets by lowering price points. A business can accomplish more with the same number of employees.
The upshot is cost savings become a complicated topic as businesses scale GenAI rollouts. For business leaders, the best approach will depend on the overall business strategy and the nuances of individual functions.
The way this will play out is case by case, area by area. Every CEO is figuring out this calculation as we speak.
Ashwin BhaveManaging director, Boston Consulting Group
"The way this will play out is case by case, area by area," said Ashwin Bhave, managing director and senior partner at Boston Consulting Group. "Every CEO is figuring out this calculation as we speak."
GenAI cuts time, but what about cost?
Software development has become a key testbed for productivity gains amid widespread adoption of GenAI coding tools. While the tools can save time, it's not clear whether they would reduce operating costs by shrinking developer teams.
Brian Greenberg, CIO at consultancy RHR International, questioned whether workforce reduction should be the goal for AI in software development. "There is definitely a time savings," said Greenberg, who uses tools such as Google Antigravity agentic integrated development environment. "But I don't look at it as headcount reduction. I look at it as augmentation. You can't show me a software development team anywhere in the world that doesn't have a two-year-long backlog." When equipped with AI tools rather than replaced by them, developers become more efficient and help draw down the backlog, he added.
Alternatively, companies that adopt GenAI strictly for the cost takeout could be disappointed. Most GenAI productivity wins, including those in software development, don't become operating cost savings, said Nate Suda, a vice president analyst at Gartner. "The productivity lift is not as high as most folks would anticipate," he explained, pointing to the 30% to 60% improvement cited by various sources.
In contrast, Gartner's research puts the typical productivity increase in the 8% to 12% range, Suda noted. "The problem with a productivity gain of 12% is that it doesn't translate into a headcount reduction, which would be the primary way for cost reduction [in] software development," he said.
While the pace of GenAI adoption has been rapid, Bhave acknowledged, "what we haven't seen is a reduction in the cost of software development just yet." Tool use, however, has lowered the demand for engineers, he said, adding that GenAI has increased output across the software development field, including requirements creation, UI/UX design, configuration testing and documentation -- as well as coding.
AI layoffs or AI washing?
AI is getting plenty of attention lately as a corporate cost-savings tactic.
Jack Dorsey, the CEO of Block, which offers Square and other financial technology products, announced in February that the company will shed close to half its employees. In a letter to shareholders, Dorsey wrote that "intelligence tools" let a smaller team "do more and do it better."
Other companies have also cited AI as a factor in layoffs, but there's another interpretation of recent staff reductions. Sam Altman, CEO of OpenAI, said some companies are "AI washing," meaning that they blame AI for cost-cutting measures they were planning to pursue anyway. His comments at India AI Impact Summit 2026 came just days before Block's layoff announcement.
GenAI reduces procurement outlays
Business leaders could find definitive cost savings in procurement. Suda cited outsourcing as a significant GenAI cost-saver for CIOs, although he noted that other business function leaders can also benefit.
In this case, AI's effects are indirect. CIOs are aware that outsourcing vendors use GenAI to reduce service-delivery costs, and they're negotiating discounts accordingly. "We're seeing CIOs, in particular, phoning up their outsourcers and getting anywhere between a 5% and 30% reduction in price off their outsourcing contracts," Suda said.
We are seeing price reductions mid-contract because of the presence of AI in the market. It increases the negotiating leverage for the CIOs.
Nate SudaVice president analyst, Gartner
IT and business process outsourcing contracts are multiyear, fixed-price arrangements, and negotiations normally take place at the end of a contract's term. But AI-influenced negotiations are surfacing in the middle of a contract, Suda said. "We are seeing price reductions mid-contract because of the presence of AI in the market," he explained. "It increases the negotiating leverage for the CIOs. [Vendors] are decreasing their prices. They know that if they don't, one of their competitors may."
A more direct cost-avoidance strategy involves using GenAI to reduce the need for hiring external service providers on a time-and-materials basis. That could include procuring anything from IT services to tax advisory assistance. Businesses are asking if their in-house personnel can use AI to perform the types of work they would traditionally parcel out to a contractor, Suda said. When the answer is yes, they can "incrementally decrease the work that's being done by a third party," he noted.
CIOs can also use GenAI-enhanced sales and accounts payable forecasting to free up cash and cut interest expenses, Suda added. Improvements in forecasting accuracy help CIOs reduce working capital needs and reliance on revolving credit.
Weighing operating cost vs. unit cost
Using GenAI-infused bots to field questions can provide customer service cost savings, Suda said. GenAI can help deflect inquiries that would usually arrive at the lowest level of customer support. In addition, enterprises are considering using agentic AI to change their structural cost-to-serve expenses.
Businesses cite this savings potential during earnings calls, Suda noted. But these discussions often focus on operational proof points rather than financial savings, he added. Companies express value as throughput -- service performance improvements -- rather than as a cost-out scenario. "It's a bit of a mixed story there," Suda said.
That variability could stem from differing motivations. Whether business leaders prioritize margin protection or market growth will influence their decision to cut operating expenses or unit costs. Also influencing their decisions is whether a particular function is seen as core to the business or more peripheral.
In areas companies deem critical to their future, they would rather reduce the cost per unit than eliminate positions, Bhave said. A biopharmaceutical company, for example, might focus its AI efforts on the cost per molecule or cost per drug, with the goal of increasing the number of drugs entering the market, Bhave conjectured. The premise isn't to cut the research budget or the roster of scientists. "The sole point is speed and output, as opposed to R&D costs going down," he noted.
A wealth management firm, meanwhile, could use GenAI to reduce the unit cost of service delivery, making it more affordable. A lower price point could help such a business extend personalized wealth advice to lower net-worth customers. In this case, GenAI would help an organization boost the number of customers its wealth advisors can support, rather than reduce the number of wealth advisors, Bhave said.
On the other hand, businesses can use GenAI to take the cost out of noncore functions. If marketing is a noncore function, a business can use AI to automate content creation, shrink their marketing spend and reinvest the savings in products or customer discounts, Bhave reasoned. Enterprises will take both approaches as they expand their GenAI use, he added.
In the end, results will depend on where and how businesses look for savings.
John Moore is a freelance writer who has covered business and technology topics for 40 years. He focuses on enterprise IT strategy, AI adoption, data management and partner ecosystems.
