Accounts payable AI glossary

AP AI · Accounts payable AI · AI-first AP

A set of AI capabilities, extraction, classification, validation, and routing, applied to the supplier-invoice processing cycle.

Traditionally, accounts payable meant reading every invoice, typing it in, validating tax, routing it for approval, and posting the journal entry. AI inverts that flow: the invoice arrives, the system understands it, classifies it, identifies the supplier, and forwards it to the correct approver without anyone keying a single number. The concept isn't OCR plus manual steps, it's an agent that makes accounting decisions. Read the full pillar: AI in accounting in 2026 →

OCR vs AI

OCR (Optical Character Recognition) extracts text from an image. AI understands that text, places it in context, and makes decisions.

OCR answers "what does this image say?". AI answers "what should I do with this invoice?". Conflating them is why so many products advertise "AI" when what they actually have is improved OCR or a chat assistant. If the system doesn't classify the GL account against your chart of accounts, doesn't detect duplicates, doesn't decide approvals, it's not AI, it's OCR. Read the full breakdown: OCR vs AI in accounts payable →

3-way match · Three-way match · Triple reconciliation

Automatic comparison of three documents, purchase order, delivery note, and supplier invoice, before approving payment. The invoice clears only if all three agree.

It's the global standard for procure-to-pay financial control. Manually: you open three documents on three screens and cross-check numbers. Automated: the system does it on receipt. If the three don't match, it escalates to the right human with the discrepancy already identified.

Intelligent invoice capture

The process by which a system receives an invoice (via email, PDF, scanned paper, XML, or EDI) and extracts the key data automatically without human intervention.

It goes beyond a simple "upload invoice" interface: it identifies the issuer, extracts line-level detail, computes tax correctly (including reverse charge and cross-border), and verifies math coherence. Good intelligent capture works directly from the inbox, no one needs to upload anything.

Touchless AP · Lights-out AP · No-touch invoice processing

End-to-end processing of an invoice, from receipt to journal entry to payment, with zero manual intervention when the invoice satisfies your configured rules.

A "touchless" invoice flows in, gets approved by rule, gets posted, gets paid, without anyone typing a number. Humans only handle exceptions: unusual amounts, new suppliers, reconciliation discrepancies. It's the highest level of AP automation.

Automated journal entry

A journal entry generated by software applying chart-of-accounts rules without human intervention, based on extracted invoice data.

Traditionally an accountant or bookkeeper does it: read invoice, decide account, write entry. Automated: the system classifies the invoice, maps the chart of accounts, and proposes (or posts) the entry ready for review. Human review concentrates on exceptions, not data entry. More context in the pillar guide: what AI in accounting automates and what it doesn't.

Reverse charge · Inversión del sujeto pasivo (ES) · Inversione contabile (IT)

A tax mechanism where the buyer (not the seller) declares and pays the tax. Applies to cross-border operations, imports, and certain B2B services.

Instead of paying tax to the supplier, you record it directly in your own tax return. If you misapply it, or fail to apply it when you should, there's a penalty. It's one of the most frequent inspection findings, especially on cross-border operations.

Bank reconciliation

The process of cross-checking bank movements against recorded invoices to confirm which invoices have been paid.

Manually: you open the bank statement, walk through invoice by invoice, mark them paid. Automated: the system matches by amount, date, and concept, and only escalates conflicts. Traditional reconciliation breaks in three cases: partial payments, batched payments, and transfer fees.

AI auditability · Decision traceability · Explainable AI

The capacity of an AI system to show, for every decision it makes, why it made it and from what data.

An AI system without auditability is a black box. If you can't explain why the AI classified an invoice as utilities or applied reverse charge, you can't defend it under a tax inspection. Auditability is what separates AI you can use in accounting from AI you can use in a demo. Go deeper in the AI in accounting pillar.