When the time comes for tech leaders to submit their budgets for the upcoming year, many of them will be in for a bad time.
After years of runaway spending, organizations are finally getting — or trying to get — their AI consumption costs under control for 2027. McKinsey recently found that around 93% of organizations are overspending on their AI allocations. Meanwhile, IT budgets are growing about 3.7% YoY, while AI spending is growing roughly 13% over the same period.
Those numbers don’t add up, and organizations and their C-suites are quickly catching on. As a result, tech leaders aren’t likely to receive the same blank check for AI solutions they’ve gotten in recent years. Instead, they may face tough questions around justifying their current AI spend — and how they plan to pay for even more in the year ahead.
Showing ROI from AI Will Be Key — But It Won’t Be Enough
At the outset of the AI boom, many companies took an “it costs what it costs” attitude toward AI. Many of these organizations were so concerned about missing out on any competitive edge AI could offer that they funded and pursued even relatively fringe AI use cases, often with little — or nothing — to show in return.
But as the calendar turns over yet again, many CFOs’ willingness to indulge in runaway spending on any and every AI solution has largely vanished. Instead of hand-waving AI budgets as they’ve done in the past, tech leaders should prepare for CFOs to ask for two things:
- Proof of ROI in the form of cost savings, labor reduction, faster speed to market, or any number of potential positive outcomes
- Which expenses they plan to reduce in order to free up AI budget
The first request is the easy (or, at least, easier) one to answer. Around 60% of tech leaders told Eliassen they’ve already seen positive ROI from their AI investments, so most should be able to find defensible numbers to show in that regard. The second question, however, is much harder to answer. Ben Murray of TheSaaSCFO.com sums it up nicely:
ETR’s data shows AI’s share of IT budgets rising from 12.1% to 14.2% in a single year, with enterprise AI spend growing around 13%. You cannot grow one line at 13% inside a budget growing at 3.6% without squeezing everything else. That is reallocation, not addition.
This “Great Reallocation,” as Murray calls it, is already forcing tech leaders to find ways to shave budget in order to continue funding their AI spending:
The Redpoint Ventures survey of 141 CIOs is where the reallocation becomes concrete. 45% say their AI budgets are coming straight out of existing software line items, not new money. 54% are running active vendor consolidation programs. Only 3% expect AI to lead to more vendors overall. Read those three numbers together, and you have the whole story. AI is not opening new budget. It is forcing a harsh re-ranking of what vendors stay in the stack.
How to Prepare for the Upcoming Cost-Cutting Discussion
With CFOs almost guaranteed to demand cost-cutting measures in order to pay AI vendors’ ever-increasing bills, tech leaders should consider the following steps well in advance of any budget discussions.
Perform a Formal SaaS Audit
Auditing your current SaaS spending is an obvious place to start, but many tech leaders may be shocked at what they discover once they get into the numbers. Zylo’s 2026 SaaS Management Index found that the average organization wastes around $19.8 million per year on unused or underutilized SaaS licenses.
While smaller organizations shouldn't expect to unearth $20 million in wasted SaaS budget, they will invariably uncover some waste, and trimming that fat — or, at least, having a concrete plan to trim it — will be crucial during upcoming budgeting discussions.
Make "Shadow AI" Its Own Audit Category
Shadow AI is still a factor at most organizations, and it may be costing yours more than you think. Personal AI subscriptions often get expensed as individual line items, and since many cost around $20-30 a month, they may be easy to miss in the regular expense-approval process.
Though these charges are small, an audit of personal “shadow AI” subscriptions can still deliver substantial savings: 100 users expensing $25 AI charges each month adds up to $30,000 in a calendar year — more than a drop in the budgeting bucket.
Stop Paying the “AI Tax”
SaaS vendors now bundle AI add-ons into higher pricing tiers, regardless of whether teams use them. In fact, procurement software provider Tropic notes that current AI pricing uplift sits between 20-47%, compared with the 3-9% pricing uplift typical of annual SaaS renewals. Any tech leader who’s had a SaaS renewal conversation recently has probably faced this “AI tax,” and while it may be common, it isn’t unavoidable.
When performing a SaaS audit, identify any vendors that levied an AI tax, and plan to renegotiate it during the next renewal cycle. Coming to the table in the next budget meeting with a plan for minimizing future AI tax charges can certainly buy some goodwill from the CFO — even if it can’t buy an immediate boost to your budget.
Additional Considerations for AI Budget Talks
ROI is great, and cost savings will be crucial, but tech leaders may want to consider other factors before submitting their budget requests.
First, AI pricing is always changing, and it’s likely to change even more substantially in the year ahead as some leading providers plan to pivot to “outcome-based pricing,” or charging only when the AI actually delivers.
This sounds good in theory. But in a world where solutions from multiple AI providers interact, and where the definition of “success” can vary from one task to another, this opens a legal and contractual can of worms most companies — and budgets — aren’t prepared to handle. These include challenges around liability, attribution, interoperability, and even data security. Coming to the table with a contingency plan for when/if key vendors make this switch can help prevent a budget crisis down the line.
Next, be aware that boards and CFOs have grown skeptical of "efficiency" and "productivity" as stand-alone justifications for additional AI spend. In Futurum's 2026 survey of IT decision-makers, the share of leaders citing direct financial impact as the primary means of ROI measurement — combining revenue growth and profitability — nearly doubled. It also replaced productivity gains as the primary way to measure the ROI of AI solutions.
Reframing ROI around P&L impact, rather than just productivity alone, may be a game-changer this budget season. Eliassen’s AI Value Navigator can help in this regard. It enables leaders to model their AI spend and see potential ROI impacts, including productivity boosts and revenue growth. It also provides clear benchmarks for how your organization’s AI operations compare to your industry peers.
Takeaways for Tech Leaders
For tech leaders accustomed to free-spending AI budgets, with no mechanisms to measure or control costs, this will likely be an especially challenging budget season.
AI is no longer experimental technology, and CFOs aren’t willing to keep issuing blank checks for additional AI capabilities. Instead, they’re asking tech leaders to provide hard numbers to justify their requests for additional AI budgets. At the same time, IT budgets aren’t keeping up with demand for additional AI dollars, so something, somewhere, has to give.
However, tech leaders who anticipate these conversations and take cost-saving steps beforehand may find budget discussions much less difficult:
- Achieving cost savings by offloading unnecessary SaaS spend and eliminating shadow AI costs is a vital first step, as will having a plan to renegotiate the “AI Tax” so common in today’s SaaS contracts.
- Planning for major changes in AI pricing can help stave off a future budget crunch.
- Demonstrating real revenue impact — not just productivity gains — from AI solutions will go a long way toward convincing CFOs to open their checkbooks this budget season.
Get a risk-adjusted ROI forecast for your AI implementations with our AI Value Navigator today, and visit our resources page today to discover more actionable insights on AI governance, cost controls, and more.
