“Modernizing payments means buying a better checkout.” That is the assumption most teams still budget around, and the latest data says it is wrong: the money is moving to the unglamorous middle of the stack — reconciliation, identity and automation — not the pay button. New PYMNTS Intelligence research, produced with Plaid, shows firms are lining their near-term spending up behind matching payments to invoices and verifying who is on the other end of a transaction, and it reframes what “payments modernization” actually is in 2026.

Reconciliation leads the near-term list

The single most-planned capability is reconciliation automation: 70% of firms plan to adopt or expand systems that match and clear payments against invoices, ledgers and bank statements within 12 months, with another 28% planning it for later — leaving only 2% not actively pursuing it. In other words, the highest-priority payments investment in the market is not accepting money; it is proving, line by line, that the money received matches what was owed.

Identity and KYC are close behind

Identity verification and know-your-customer (KYC) automation come next, with 65% planning adoption or expansion inside a year. Include longer-term plans and the figure reaches 90% — 25% more have it scheduled for later, 9% are considering it without a timetable, and just 1% rule it out. That makes identity one of the most broadly planned capabilities in the entire survey.

AI is on the agenda, but it doesn’t lead

AI-based fraud detection lands at 59% for near-term adoption — tied exactly with secure bank connectivity and open banking. AI matters, but the data punctures the idea that it is the whole story: a further 23% plan AI fraud tooling for later and 15% are weighing it without a schedule. And modernization is additive, not a teardown — 51% still plan to spend on existing legacy infrastructure even as they layer new capabilities on top.

Why the “boring” layer is the real story

Read together, the numbers describe a shift in where payments value is being built. The front end — the card field, the wallet button — is largely solved. The hard, high-priority work has moved to the ledger: can you automatically reconcile every inbound payment against the invoice that authorized it, and can you trust the identity behind it? Those are accounting and trust problems, not checkout problems.

It is a useful lens for anyone choosing commerce infrastructure. A platform built with the invoice as its centre of gravity — where every payment settles against a specific, timestamped line item and an event-based engine keeps assets and obligations in balance — treats reconciliation as a native property rather than a bolt-on integration. The self-hosted, source-available VBWD platform is built exactly that way: invoice-centric with a provider-agnostic payment core, role-based access for the identity and KYC seam, and reconciliation that falls out of the model instead of being stitched across it. For teams whose 2026 roadmap looks like this survey, that is worth a look — the documentation and the public SDK on GitHub show how it fits together.

Written for Red Robot with AI assistance and human editing. Based on reporting by PYMNTS.