Scaling AI with Vision
Why agentic workflows will redefine forecasting, close, and compliance
At the risk of appearing to suffer from split personalities, I’d like to spend some time this week looking at the other side of the coin when it comes to the promise and future of AI in the workplace …
Last week we looked at the MIT Sloan study that showed 95% of enterprise generative AI pilots fail to deliver measurable business impact. The message was clear: without strategy and governance, AI projects are destined for “pilot purgatory.”
But failure isn’t the whole story. In fact, it’s the exception when AI is deployed with foresight, planning, and vision. This week, let’s flip the lens and explore the upside: what finance teams are already achieving with generative AI and where this technology is taking us next.
Adoption Is Accelerating
Finance leaders aren’t waiting on the sidelines. Nearly all CFOs (96%) say they are prioritizing AI integration, and 79% plan to increase spending on AI in 2025 (CFO Dive – May 2025). Gartner projects global AI spending will reach $644 billion in 2025, up 76% from 2024 (Gartner via CFO Dive).
Yes, adoption is still in its early stages. Only 4% of finance leaders say their organizations are “maturing and scaling” GenAI (CFO.com – Apr 2025). But that gap between intention and maturity is exactly where the opportunity lies.
CFOs overwhelmingly identify AI as the #1 force reshaping finance over the next five years (CFO.com – Jul 2025). This optimism reflects measurable returns, not speculation.
The ROI Is Already Real
Properly deployed generative AI in finance has moved past “interesting experiments” into measurable impact. In mid-2025, nine in ten CFOs said they were seeing very positive ROI from GenAI projects, which is a dramatic shift from just 27% reporting the same a year earlier (PYMNTS – Sep 2025).
Finance professionals estimate that AI tools are already saving them five hours per week per employee, worth about $19,000 per year per person (CFO.com – Jul 2025). Processes that used to take days (e.g. variance analysis, management reporting, reconciliations) are now compressed into hours. And the Hackett Group forecasts 44% productivity gains across SG&A functions once GenAI scales (Hackett Group – 2025).
Case Studies: Finance at Work With AI
Early adopters are proving what’s possible:
EY’s Agentic FP&A Dashboard: EY built a forecasting tool that deploys multiple “silent analyst” agents to generate real-time scenarios. A manager can ask, “What happens if GDP drops 1%?” and get an instant modeled answer. As one client put it, “I have an army of silent FP&A analysts now” (CFO.com – Jun 2025).
KPMG’s Clara Audit Platform: Clara automates tedious audit tasks like expense validation (ingesting documents, extracting data, and drafting workpapers). Auditors focus on judgment calls instead of manual sampling, resulting in faster audits and higher quality outcomes (CFO.com – May 2025).
Financial Close Co-Pilots: Companies are piloting GenAI to flag anomalies in journal entries and draft narrative commentary for reports. Roughly one-quarter of finance leaders say they are testing AI in general accounting and close activities, with another 20% piloting accounts payable automation (CFO.com – Apr 2025).
These examples are already in motion, proving how GenAI can augment finance work.
Why Projects Succeed
As we covered last week, too many pilots fail from lack of strategy or poor data hygiene. But the successful projects share common traits:
A clear strategy tied to finance goals, not one-off experiments.
Early use cases with clear ROI: forecasting, reconciliations, reporting.
Strong governance and controls, treating AI like any other financial system.
Upskilled teams who know how to use and review AI outputs.
Cross-functional collaboration between finance, IT, and compliance.
Organizations with tailored AI roadmaps are already 3.5× more likely to report ROI than those dabbling without a plan (CFO.com – Jul 2025). The technology works when the approach is right.
The Road Ahead
Regulators are signaling that AI is becoming part of the fabric of finance. The PCAOB has made clear that auditors can use AI, but “quality control cannot be replaced by AI” (PCAOB Spotlight – Jul 2024). The SEC is holding roundtables on AI in financial reporting and reminding firms that existing rules still apply: if AI touches financial disclosures, accuracy and controls remain non-negotiable (SEC – Mar 2025). And the EU AI Act is classifying many finance applications (like credit scoring and fraud detection) as “high-risk,” requiring more transparency and oversight (The Treasurer – Mar 2025).
For finance leaders, this is good news. It means AI is here to stay, and finance can lead by showing how to use it responsibly so that inputs and outputs are auditable, explainable, and regulator-ready.
The real story isn’t that AI pilots fail. It’s that when done right, AI transforms finance. We’re already seeing it in FP&A, audit, and the close. The CFOs who combine vision with governance will unlock faster closes, sharper forecasts, and more resilient compliance.
The long-term promise of AI’s applications in finance warrant investing time and effort in deploying it today. As someone who’s been working with AI and ML in finance for almost a decade, I may be among the converted, but I can see in the next five to 10 years, finance teams won’t just be using AI as a “copilot.”
They’ll orchestrate networks of autonomous agents running in the background. Forecasts will update themselves continuously as actuals flow in, close cycles will shrink from weeks to hours, and board packs will be generated on demand with commentary drawn from live data. Instead of building models from scratch, analysts will supervise AI agents that simulate dozens of business scenarios in parallel, stress-testing assumptions in real time.
In this world, the finance function shifts from backward-looking reporting to a continuous strategy engine where the CFO manages a portfolio of AI agents as carefully as they manage capital today.
The next chapter isn’t about experimenting with AI. It’s about scaling it responsibly. And finance, with its culture of control and trust, is positioned to lead the way.
From Insight to Action
What we’ve seen here is the upside: AI isn’t just theory, it’s already reshaping finance. But the bigger question is how you scale it without losing control.
In the Pro version, we dig into:
ROI in practice – with a calculator you can apply to your own team.
Agentic workflows mapped step by step – from close-to-forecast handoffs to board pack prep.
Governance in action – how to make every AI-assisted process audit-ready.