- 60% of businesses still use checks for B2B transactions (2024)
- $15 trillion in nominal value moves through paper checks annually
- Processing a single paper check costs $4–$20 per transaction
Experts would likely conclude that while AI-driven solutions like Ordway's are bridging critical gaps in outdated payment systems, the persistence of paper checks highlights deep-rooted inefficiencies in B2B transactions that require both technological and behavioral shifts.
The Analog Paradox: AI Meets the Stubborn Persistence of Paper Checks
WASHINGTON, D.C. — November 17, 2026 — We live in an era where artificial intelligence can generate complex software code, diagnose rare diseases, and pilot autonomous vehicles. Yet, step inside the accounting department of a mid-sized American corporation, and you will find highly educated professionals engaged in a distinctly 19th-century activity: deciphering handwritten paper checks.
Today, Ordway, a Washington-based finance platform known for its focus on complex business models, announced a new capability for its AI-driven cash application agent. The system can now parse check images and detailed bank lockbox reports to automatically reconcile lump-sum paper check deposits against open customer invoices.
At first glance, it is a mundane software update. But look closer, and it reveals a fascinating, almost absurd paradox at the heart of the modern economy: we are deploying the most advanced technology in human history to patch the inefficiencies of an analog payment method that refuses to die.
The Stubborn Survival of the Paper Check
To understand why a company like Ordway—which recently secured $20 million to accelerate its AI product development—is focusing on paper checks, one must look at the glaring gap between the utopian visions of Silicon Valley and the messy reality of American back-offices.
Despite a decades-long push toward digital transformation, business-to-business (B2B) payments remain stubbornly tethered to paper. Recent industry benchmarking data shows that as recently as 2024, nearly 60 percent of businesses still used checks for B2B transactions. Among small businesses, that number hovered near 80 percent. While the sheer volume of checks has declined over the last decade, the nominal value moving through this paper pipeline has remained relatively stable at a staggering $15 trillion.
"It's amazing that in today's AI-powered economy, paper checks remain such a meaningful percentage of B2B payment volume and the pain it causes is real," said Sameer Gulati, CEO and Founder of Ordway. "Our customers spend days every month trying to reconcile lockbox deposits with open invoices."
Why do they persist? For many legacy businesses, particularly in construction, utilities, and professional services, checks offer a perceived sense of control over cash flow. More often, it is simply institutional inertia. Upgrading accounts payable systems requires capital and IT resources that many companies prefer to spend elsewhere. The result is a hidden tax on the economy. Processing a single paper check costs a receiving business anywhere from $4 to $20 when factoring in bank fees, labor, and postage. With the Federal Reserve projecting a 5.4 percent average increase in check-service prices, the financial burden of this analog habit is only growing.
The Month-End Drudgery of Lockbox Reconciliation
For decades, banks have offered "lockbox" services as a band-aid for the paper problem. Customers mail their checks directly to a bank-managed post office box. The bank opens the mail, deposits the checks, and sends the company a daily data file and a lump-sum deposit.
While lockboxes accelerate the physical clearing of funds, they create a massive data abstraction nightmare for accounts receivable (AR) teams. A single $50,000 lockbox deposit might represent forty different checks paying off seventy different invoices.
This is where the human toll of inefficient systems becomes glaringly apparent. Every month, AR clerks and treasury analysts are subjected to the soul-crushing drudgery of "cash application"—the process of matching incoming cash to outgoing bills. They stare at dual monitors, comparing blurry check scans against endless spreadsheets, playing a high-stakes game of memory and deduction.
The discrepancies are maddening. A customer might consolidate payments for five invoices onto one check but forget to include the memo. Another might short-pay an invoice due to a dispute over a damaged shipment. Overpayments occur. Payor names on the check often fail to match the customer account record in the enterprise resource planning (ERP) system because a subsidiary paid on behalf of a parent company.
Ordway's updated AI agent is designed specifically for these "no match" scenarios. By utilizing probabilistic and deterministic matching logic, the software analyzes the check images and remittance data, cross-referencing it with historical customer behavior to accurately apply the funds. It effectively automates the intuition of a seasoned accounting clerk.
Keeping the AI on a Short Leash
However, the introduction of AI into corporate ledgers is not without its anxieties. In the realm of corporate accounting, where regulatory compliance and audit trails are paramount, a "black box" algorithm is a non-starter. Finance departments demand absolute explainability.
Recognizing this, Ordway designed its AI agent to operate in a "supervised mode." Rather than autonomously writing entries to the general ledger without oversight, the system presents its proposed matches and shares the details of its reasoning with the accounting team.
"We are dealing with the lifeblood of the company—its cash," noted one corporate controller at a mid-market manufacturing firm who oversees a team of AR specialists. "I cannot tell an auditor that a million-dollar discrepancy exists because the computer thought it was a good idea. We need to see the math. We need the ability to pull the emergency brake."
This human-in-the-loop approach allows finance professionals to oversee the process and execute manual overrides when the AI encounters an unprecedented edge case. It is a critical design choice that acknowledges the limitations of machine learning when faced with the infinite variability of human error.
Ordway is not alone in this space. Incumbents like HighRadius, Billtrust, and BlackLine have spent years developing AI-powered cash application software, boasting high touchless match rates. Billtrust, for instance, utilizes a confidence-based matching engine to handle messy data, while HighRadius targets 90 percent automation.
What differentiates Ordway is its broader ecosystem. The platform is tailored for technology-centric companies navigating complex monetization strategies—subscriptions, usage-based billing, and outcome-based pricing with intricate contract structures. By integrating lockbox reconciliation directly into a system that already handles recurring billing and revenue recognition, Ordway allows finance organizations to manage the entire quote-to-cash lifecycle within a single, unified environment.
Bridging the Gap Between How We Work and How We Should
The development of AI to read paper checks is a profound commentary on the nature of technological progress. We often imagine innovation as a clean break from the past, a sudden leap into a frictionless digital future. The reality is far more compromised. Progress usually looks like an advanced algorithm learning to read the messy handwriting of a clerk who refuses to stop using a checkbook.
By addressing this unglamorous reality, companies like Ordway are doing vital work. They are not just reducing Days Sales Outstanding or cutting administrative overhead; they are liberating human workers from robotic tasks. The true promise of artificial intelligence in the workplace is not the elimination of human oversight, but the elevation of human labor.
When an accounts receivable manager no longer has to spend three days a month squinting at misaligned check scans, they can finally focus on the work that actually matters: resolving disputes, analyzing credit risks, and understanding the financial health of the business. In the end, the most powerful application of AI might just be its ability to handle the analog baggage we refuse to leave behind.
Topics & Related
Fintech
📝 This article is still being updated
Are you a relevant expert who could contribute your opinion or insights to this article? We'd love to hear from you. We will give you full credit for your contribution.
Contribute Your Expertise →