That is one central takeaway from the December 2025 “Invoice-to-Pay Automation Tracker Series,” a PYMNTS Intelligence report, which examines how enterprises are using AI to modernize AP while still relying heavily on manual processes.
AI adoption in AP is widespread, according to the survey. But its most valuable contribution may not be simply accelerating the processing of invoices and reducing errors through automation, but in its ability to unify fragmented invoice, approval and payment data and convert it into usable insight that supports better financial decisions across the organization.
The survey found that manual invoice processing remains dominant, even among companies that describe themselves as digitally mature. These manual steps hide spending patterns, weaken forecasting accuracy and make it harder to detect anomalies that could indicate errors or fraud early.
AI addresses those gaps by standardizing data fields, improving accuracy and creating a single view of obligations and supplier activity. When those data are visible and reliable, finance teams can move beyond transaction processing toward analysis that informs working capital strategy, payment timing and supplier management.
Crucially, the findings suggests that many enterprises have only scratched the surface of what AI can reveal. AI systems can highlight drift in spending categories, flag suppliers whose invoicing behavior signals risk, and model the downstream impact of different payment strategies. These insights were often buried in disconnected systems or paper-based workflows. By making them visible, AI allows AP to function as an early warning system and a planning tool, rather than a back-office cost center.
Read the report: Smarter Spend: AI-Powered AP for Data-Based Decision-Making
Key data points from the report illustrate both progress in AI adoption and missed opportunity in its use:
- 79% of organizations that use AI report measurable performance improvements, including faster invoice processing, quicker approvals and improved employee satisfaction.
- One-third of CEOs believe cash forecasting and spend analysis are the areas that would benefit the most from AI, yet only 26% are prioritizing these use cases. Encouragingly, 82% plan to invest in AI for AP in the next 12 months, with 34% intending to prioritize its cash-flow forecasting ability.
- 72% of companies say they have adopted AI in AP within the past two years, yet only 22% report full, at-scale usage, indicating that most deployments remain limited in scope
To capitalize on the hidden signals AI can uncover, the report emphasizes the importance of trust and governance. Finance leaders remain cautious about relying on AI outputs without clear visibility into how decisions are made. Concerns about data privacy, security and accuracy continue to slow broader adoption. As a result, most organizations favor human oversight, either through review of every AI decision or by focusing human attention on exceptions.
The report concludes that AI works best as a copilot rather than a replacement. When paired with clear approval logic, transparent decision rules and defined escalation paths, AI can surface insights faster while preserving control and auditability.
PYMNTS Intelligence recommends that organizations prioritize strategic use cases such as spend analysis and cash forecasting, advance automation in phases, and build governance frameworks early. Those steps allow enterprises to move from basic workflow acceleration to intelligence-driven planning, unlocking value that has long been hidden inside accounts payable data.
