AI can make finance faster and smarter. That promise is getting louder as the conversation shifts from copilots, automation and agents toward the possibility of computational capabilities that bring us closer to superintelligence for the benefit of organizations and businesses.
For CFOs, though, the immediate question is more practical: which AI investments are creating measurable value now, and which are simply adding cost, complexity and risk?
In From Insight to Infrastructure: The Intelligent Age and a New Path Forward, I argue that AI is becoming the infrastructure underneath the business, not simply a layer on top of it. Emburse is among the 2% of comparable vendors that have embedded AI in mission-critical business processes, delivering customer value while prioritizing trust and security.
The debate around superintelligence is a useful reminder that the technology is moving faster than many operating chains of command can keep up with. We cannot manage it with old approval rules or with a blank cheque stamped AI. The opportunity is to redesign workflows to maximize benefits with less wasted effort, protecting the enterprise while allowing good decisions to move through it. Consider a travel policy focused on the lowest cost per trip. It might discourage a customer visit that could open a new market or help secure a major account. The savings are easy to measure. The potential return is harder to capture unless leadership is thinking holistically and the C-suite is engaged.
AI changes the cost conversation
Companies are paying for AI in more ways than a software licence. There are usage-based charges, data and integration work, security reviews, training, support and the time employees spend learning new tools. Because AI can be easy to buy, similar capabilities can multiply across the business before leaders have a clear view of what they are paying for or receiving. Decades of experience using AI to amplify human productivity inform how we build today. We apply AI where it delivers measurable ROI, helping teams accomplish more with existing resources and make better-informed decisions. We do not chase “tokenmaxxing.” We focus on AI that gives customers a meaningful competitive advantage.
This requires ROI discipline: an operating habit every day, every week and every month, not a once-a-year exercise. Before approving an AI investment, leaders should be able to say what work it will improve, who owns the outcome and what evidence will show that it is working. If the answer is vague, either because the workflow or the business purpose is unclear, the investment should not get a pass simply because it carries an AI label. Projects without a clear path to value waste resources that could otherwise be used to maximize returns where results are clearly identifiable and can be achieved quickly. That is the standard finance should apply as it decides how much control and how many resources each choice warrants.
Emburse provides out-of-the-box productivity gains for our customers. But the opportunity is bigger than a card or a reimbursement workflow, given the data that flows through a system of record. Finance is primed for connected, intelligent controls that help people make sound choices across travel, expense, invoices and payments, while giving leaders a clearer picture of the cost and value of those choices. The best finance software should be more than a rear-view mirror. It should help the business decide what to do next to optimize spend. Productivity remains essential because human capital is one of the largest spend drivers in a business. Hosting costs are also a major driver for many businesses. Add travel, and you can see how it all fits together and why the solutions we provide are essential for our customers.
Control should reflect the risk
Finance rules exist for good reasons. They protect company resources, uphold legal and regulatory obligations and make decisions more accountable. That discipline is essential for complex enterprises, but it can become a hindrance when it is not applied well. Complexity is high, and so is the cost of getting it wrong. Yet many policies were written before information was available in real-time, so organizations built human checkpoints into even routine decisions. The people involved in routine decisions do not go away. Instead, they can do more and focus their insight where human judgment is needed. When approvers approve everything, the workload shifts to auditors, who are costly and overworked. AI can help by changing the initial behavior, supporting approvers or replacing that step, and helping auditors focus where attention is needed rather than trying to audit everything.
Today, real-time policy guidance and earlier insight into spending patterns and anomalies can help finance direct attention where it counts. Routine, low-risk expenses may move with little intervention. Unusual or high-stakes decisions deserve greater context and human scrutiny. The principle is to match the level of oversight to the decision, rather than make everyone pass through the same approval process.
Good control also has to work for the people following the rules. Employees are more likely to comply when guidance is clear and the right choice is easy to make. A prompt before a purchase helps more than an exception discovered weeks later. Convenience, designed responsibly, can strengthen compliance and leave employees better able to do their jobs.
This is a more useful definition of control: help people make sound decisions, identify real risks early and keep accountability where it belongs. It gives finance a way to support growth without confusing speed with a lack of discipline.
Investment rules should follow business priorities
The same reasoning applies to what companies choose to fund. A purchase can satisfy every approval and still deliver little value. Meanwhile, an initiative that could open a new market or improve a customer experience may lose momentum because it falls outside a rigid budget category. The CFO should test both business value and risk.
Our research on AI spending shows how easily an incentive can pull decisions off course. In a survey of 1,500 corporate decision-makers in the U.S. and U.K., 58% said AI-related software was easier to approve than other software, and 62% reported connecting at least one purchase to an AI initiative to secure funding. These answers suggest that the label on an investment can sometimes carry more weight than its business case. In an environment where AI costs can grow quickly, that is a habit worth challenging.
That calls for a consistent test. Before approving an investment, agree on the problem it should solve, the full cost of getting it into use, who owns the outcome and what evidence would demonstrate value. Keep reviewing those commitments after the contract is signed. Finance cannot do this alone: business leaders understand customers and operations, technology teams understand systems and security, and procurement understands suppliers and commercial terms. The CFO’s contribution is to connect those perspectives and keep spending aligned with strategic priorities.
Give people room to make better decisions
Changing finance rules also changes what we ask of people. When automation handles routine checks, finance professionals can spend more time on exceptions, advice and emerging risks. Managers who receive greater discretion need clear information, training and boundaries.At Emburse, we see this in small, familiar moments: an employee unsure what a policy allows, a manager approving without context, or an analyst chasing information that should already be available. AI fluency needs to reach beyond the executive team. Employees must know when to use a recommendation and when to question it. Finance teams need the authority to turn better information into smarter decisions. We see that balance of speed, control and judgment in our partnership with the TGR Haas F1 Team, where fast decisions work only when everyone understands the guardrails.
Start with one rule that works. Look at a policy that is being followed exactly as written. Consider the risk it was designed to manage, the behavior it encourages and whether the result still advances the business’s priorities. Keep safeguards that protect the enterprise. Redesign requirements that make good decisions unnecessarily difficult. AI is making finance faster. The bigger opportunity for CFOs is to shape control that is rigorous where risk demands it, responsive where opportunity calls for it and clear enough for people to act with confidence.Every rule should face the same test: does it still protect the business, or is it standing in the way of the company we need to become? Your AI strategy should likewise focus on making you a better business, rather than on distracting projects that do not scale and take more time to make work than they deliver in return. You want cutting-edge vendors that help you become the best business you can be.