Many tech companies push AI innovation for clients but fail to apply it in their own finance functions, missing cost reductions of up to 40% from end-to-end agentic AI, according to EY analysis. Siloed data, unclear ownership and change management hurdles keep finance teams stuck at basic automation.
From RPA 2.0 to agentic workflows
The current state of AI in finance at many tech firms could be described as "Robotic Process Automation 2.0." AI handles parts of a process - invoice extraction, anomaly detection - but not the whole. Isolated point solutions trim costs but do not reshape the function or cut headcount significantly.
Agentic AI moves beyond single-task automation. It creates repeatable, intelligent workflows that scale across teams and geographies. Routine tasks such as data consolidation, variance checks and first-draft commentary can be taken over entirely, freeing analysts for higher-value work. "When we shift our thinking from eliminating single tasks or processes to designing collaborative outcomes, AI delivers significantly more benefits," said Adam Blaylock, EY Americas Financial Accounting Advisory Services TMT Industry and Technology Sector Leader.
Designing finance with zero human intervention in mind
A "design for zero" approach means reimagining processes to reach outcomes without human touch, then adding people only where strictly necessary. Blaylock said this starts with visualizing the future state: how finance operates when repetitive work is fully automated, what new roles and skills are needed, and which standard policies let AI function across the department. The outcome-based mindset replaces a piecemeal renovation.
Where to build and where to buy
Finance leaders often face a choice: use AI embedded in enterprise resource planning software or build internal tools. EY professionals recommend that tech firms prioritize internal development for order-to-cash (OTC) and financial planning & analysis (FP&A), because these processes are unique to each company's market position. Procure-to-pay (PTP) and record-to-report (RTR), by contrast, are more standardized and may be better served by AI features that software vendors will embed over time.
"The important thing to remember is that there is no 'one size fits all' solution and companies need to take a deep look at their internal processes and supporting tech before moving forward," said Amanda Donohue, Principal, Finance Consulting at Ernst & Young LLP. "The key is to invest in reimagining your critical and complex processes rather than waiting for someone else to reimagine them for you."
Compliance and accuracy remain non-negotiable
Tech firms demand certainty in internal and external reporting, and the rapid growth of AI leaves regulators playing catch-up. Teaming with a trusted advisor that understands controls, regulatory reporting and AI assurance can accelerate adoption. Blaylock said agentic AI, when properly designed, improves data accuracy by removing human error and freeing time for review. "We are always working with regulators to improve financial reporting processes and standards, and we believe AI has a major role to play today and in the future," he said. Many clients are moving toward a managed service model, drawing on global pools of analysts, engineers and accounting professionals with proprietary toolkits.
AI validation tools are also becoming necessary. While AI makes it easy for employees to write code and create automated processes, there is no guarantee the outputs are correct. Organizations must monitor proprietary model performance continuously.
Three hurdles slowing adoption
Finance officers grapple with three common concerns: choosing which AI tools to invest in, assessing workforce readiness and upskilling employees, and integrating AI with major enterprise software upgrades. Each of these challenges points back to the broader need to rethink the division of work between AI agents and people. For finance leaders addressing workforce readiness, structured AI training programs such as an AI Learning Path for CFOs can provide direction. Additional resources covering AI for Finance trends and case studies are also available.
Why this matters for finance professionals
Agentic AI is shifting finance work into two buckets: routine operations managed by AI with human oversight, and strategic tasks handled by people. For FP&A analysts, this means data is already gathered and analysis drafts are pre-built when they start their day - allowing them to focus on insight and business partnering. Finance teams that adopt end-to-end agentic workflows can deliver the same output with reduced headcount while improving accuracy and speed. The question is no longer whether to use AI, but how quickly to move from piecemeal experiments to a redesigned finance function, a journey Blaylock said can happen within months.
Your membership also unlocks:
