While there has been a lot of buzz about AI in recent years, today, AI is genuinely transforming all aspects of business, including the finance function.
A recent survey, The State of AI in Finance: Exploring the AI Trust Gap, completed by global finance professionals, shows that while 55% of respondents are extremely optimistic about the potential benefits of AI, 48% are still concerned about the risks of using it.
Managing the AI trust gap is imperative for financial professionals to adopt it at scale, leverage its benefits, and focus on more impactful work. To better understand the benefits of AI and how its risks can be managed, we’ve asked recent participants in our Next Gen Finance Leaders series about their first-hand experience and views.
Key Takeaways
- 55% of finance professionals are optimistic about AI’s potential, but 48% still worry about its risks, according to The State of AI in Finance survey.
- Across the Next Gen Finance Leaders interviewed, one point holds: AI still needs guardrails, human judgment, and critical thinking, especially when outputs miss the business context an experienced professional would catch.
- Trust in AI isn’t given, it’s earned, through clean data, vendor trial periods, ongoing AI training, and keeping humans in the loop on review.
Human Judgment Is Still Key
Sabrina Carrion, Global Accounts Payable Manager at Atlas HXM, is seeing the AI shift firsthand: her team is using it to work faster and focus on more meaningful work instead of repetitive tasks. She believes AI is becoming as expected in finance as automation adoption was a few years ago—though she’s quick to stress that AI-supported workflows still need human judgment, since outputs can be misleading if they don’t account for the wider business context that an experienced professional understands better.
I expect AI to take care of routine work so finance professionals can spend more time analyzing, solving problems, and partnering with the business.
Sabrina Carrion, Global Accounts Payable Manager, Atlas HXM
Guardrails Build Trust
Christian Sanford, founder of QuantFi, a full-stack strategic finance agency, is broadly positive about the benefits of AI—enough that he’s rebuilt the company around agents rather than adding additional headcount. But his trust in AI isn’t binary; it depends on whether his team has put checks and balances in place. For Sanford, the most important capabilities in any AI product are “the ability to see and review the action the AI takes” and “being able to custom-configure [tools] to automate tasks in specific ways.”
I trust the outputs where we’ve built the guardrails to verify them, and I treat the rest as suggestions rather than truth. Finance is a discipline of provenance, and any tool that hides its reasoning or its steps eventually breaks under audit.
Christian Sanford, Founder, QuantFi
Critical Thinking Can Unearth Errors
Joseph Falcao, founder of Falcon Pax, an advisory firm that supports high-growth companies, says AI isn’t reinventing the CFO role. Instead, it’s enabling finance leaders to fulfill the position’s true purpose: being forward-looking, embedded in the business, and focused on value drivers to make agile, strategic commercial decisions backed by fast, accurate data. He points to redesigning and automating processes, including payables, as some of the best use cases for AI in finance, and argues that professional experience and critical thinking remain vital to managing trust.
AI in the CFO office is only half the journey. Finance people need to understand data definitions and assumptions behind the numbers to review AI outputs. It still takes sharpened, critical thinking to know what to do with them, and to know when they are wrong.
Joseph Falcao, Founder, Falcon Pax
Trust Needs to Be Earned
Akshay Shrimanker, CPA, founder of ShayCPA, a NY-based firm specializing in venture-backed startups, is broadly optimistic about AI, particularly for supporting month-end processes and tax compliance. However, he says trust needs to be earned, rather than given. In practice, this means that the firm takes on trial periods for new AI vendors and implements a 20-hour minimum annual AI training commitment.
It’s important that the team can get comfortable using the product, understand the terms of service, and ensure the product is actually solving a problem our clients have. We want them to have exposure to a wide variety of AI-based training.
Akshay Shrimanker, CPA, Founder, ShayCPA
Don’t Treat AI as the Final Solution
Michael Winter, CFO at Lantern Community Services, a New York-based nonprofit that provides services for people at risk of homelessness, thinks AI can “provide multiple organizational benefits.” But he’s careful to frame that potential with a caveat.
AI is a tool, not a solution, so adoption should be grounded in rules and usage guidelines. It shouldn’t soften the individual’s capacity to think.
Michael Winter, CFO, Lantern Community Services
Clean Data as a Starting Point
Denise Meyer, Accounts Payable Staff Accountant at Voltus, a demand response and distributed energy resource management company, is optimistic about AI for its potential to reduce data entry errors and minimize the risk of misinterpreted data—enhancing both productivity and accuracy across financial operations. For Meyer, managing that risk starts with clean data, and the main barrier to adoption she sees is integration with existing legacy systems.
Ensuring our data is clean reduces one failure mode. From there, the remaining risks of misapplied methodology, miscalculations, and misinterpretations should be supported by random checks, parallel analysis, and human-in-the-loop validations.
Denise Meyer, Accounts Payable Staff Accountant, Voltus
Closing the AI Trust Gap
Across these conversations, one theme holds steady: trust in AI isn’t assumed, it’s earned—through clean data, clear guardrails, and human judgment that stays firmly in the loop.
Our Next Gen finance leaders aren’t waiting for AI to be flawless before putting it to work; they’re closing the trust gap deliberately, through training, review processes, and a clear sense of where AI’s role ends and their own expertise begins. That’s the real path to adopting AI at scale: not blind confidence in the technology, but the discipline to manage its risks so finance teams can capture its benefits and focus on the work that matters most.
