Closing the readiness gap.
Artificial intelligence has entered nearly every boardroom conversation, but in Finance and Accounting (F&A), the promise is still far ahead of the practice.
A poll of over 300 attendees in a recent magentIQ webinar revealed that 62 percent of finance leaders are still “researching and experimenting” with AI. Only 11.3 percent are running pilots, and fewer than 5 percent have systems in production.
Infographic: State of Agentic AI in Finance & Accounting (click image to enlarge)
That mirrors global trends. According to McKinsey, 74 percent of companies struggle to scale AI beyond pilots, while BCG reports that just 4 percent of enterprises generate real value from AI. The result is widespread frustration: ambitious projects stall, costs mount, and CFOs are left with little to show for their investments.
Ian Barkin, co-founder and co-CEO of magentIQ, put it bluntly during the session: “95 percent of AI projects fail… and what we tend to find is projects don’t fail because of the tools being used. It’s because of everything else.”
The readiness gap
That “everything else” refers to the operational realities of finance functions. Data silos, legacy systems, and unclear processes undermine even the most advanced AI. As Barkin explained, companies often lack the “substrate, the soil in which you planted those seeds.” Without clean data, documented workflows, and organizational buy-in, AI implementations resemble science projects rather than business transformations.
Kelly Dunn, who leads Finance Transformation at Georgia Pacific, echoed that sentiment: “We’re trying to resist the emphasis around solutions and just focus on what the problems are we’re trying to solve… and doing that within a disparately data landscape.”
This problem-first mindset is critical. “Fall in love with your problem, not your solution,” Dunn urged. Yet in practice, many organizations reverse the order-falling for vendor hype and rushing to install tools without first defining the business outcomes they need.
Where pilots go wrong
AI pilots often start in isolation. An invoice extraction tool here, a chatbot there. But disconnected tools rarely add up to measurable impact. “Most pilots stall because they automate tasks in isolation,” noted David Brain, magentIQ’s co-founder. “The wins we see come from outcome-focused sprints where AI agents and human operators work together, end to end.”
That distinction matters. Automation of tasks is not the same as transformation of outcomes. Automating journal entries may save minutes. Reengineering the month-end close with AI anomaly detection and predictive entries can save days-while improving accuracy and auditability.
Tariq Munir, author of Reimagine Finance, was even more direct: “Today, we are in 2025… as a base, your entire back office support or your entire transaction processing should be automated. There is no excuse for that.”
How to avoid the pitfalls
For finance leaders seeking to capture AI’s promise, three strategies stand out:
- Start with outcomes, not tools – Barkin emphasized that success requires defining the destination before picking the journey: “If you solve the problems, you create an outcome… you’re trying to pick your destination before you pick the tools and the journey you take.”
- Embrace experimentation, but expect missteps – As Dunn admitted, “We absolutely picked the wrong first use case for our first experiment, but we learned a ton along the way.” The lesson: every attempt, even failed ones, moves teams closer to scalable solutions.
- Combine AI with human oversight – MIT Technology Review has warned against blind faith in AI agents, noting that “these systems still struggle with long-term memory, general planning, and making reliable choices in unfamiliar contexts.” The solution is orchestration: AI for speed and scale, humans for context, compliance, and judgment.
The path forward
The future of AI in Finance and Accounting is not about replacing people with machines. It is about building what magentIQ calls a multi-agent workforce-AI agents and human specialists working side by side, each deployed to the right task at the right time.
That approach delivers immediate benefits. Barkin highlighted that by combining global talent with AI, magentIQ is “able to deliver 40 plus percent savings at the very beginning of a journey.” Munir added that discrete projects can show ROI in as little as three months, while more complex processes may take six to nine months.
The message for CFOs and controllers is clear: AI adoption does not need to be a moonshot. It can be a series of well-defined milestones. Start with the right problem. Run a governed, outcome-driven experiment. Learn fast, then scale.
Key takeaway for F&A leaders
AI in Finance & Accounting fails when it chases hype, tools, or isolated pilots. It succeeds when leaders anchor efforts in outcomes, embrace experimentation, and orchestrate AI with people. The winners will not be those who deploy the most agents, but those who deliver measurable results.
Related content
What Cognitive Offloading Is, and How to Keep Your Team Sharp While Using AI
Cognitive offloading is the delegation of mental work to something outside your own head, and generative AI extends it from arithmetic and directions to analysis and judgment. What the research establishes, what it does not, and five practices that keep your team sharp.
WHITEPAPER: Designing the modern Accounting Firm with Agentic AI
Preparing your firm for the shift to an agentic operating model.
WHITEPAPER: The Map Before The Machine
How mapping the work before choosing the technology enables leaders to build an agentic operating model that delivers measurable returns.
Why Teams Stop Trusting AI: 5 Ways to Earn It Back
McKinsey found that 80% of respondents report personal productivity gains from AI, while 37% can point to any EBIT impact. Part of that distance is trust. Five patterns that break a team’s confidence in AI at work, and the design decision that rebuilds each one.