KYB Is the Bottleneck for B2B Stablecoin Payments

Persona’s July 2026 report Into B2B Stablecoins found 76% of fintechs experienced document fraud in the prior year. The report identifies Know Your Business (KYB) checks as the gating control for B2B stablecoin payments because on‑chain settlements cannot be reversed once value is moved.

The report, published in July 2026, examines verification and fraud trends affecting stablecoin platforms and card issuance programs. It defines KYB as verification of an organization’s legal entity, its registered representatives and its ultimate beneficial owners (UBOs). KYB requires confirming legal existence and good standing, identifying controllers, and screening relevant parties against sanctions and adverse‑media lists. The report contrasts KYB with Know Your Customer (KYC), which verifies an individual.

Persona identifies three structural challenges that make KYB harder for stablecoin platforms. First, stablecoin tokens can circulate across markets instantly, forcing platforms to query many business registries at once. Registries vary in data quality, naming conventions and access rules.

Second, regulation is fragmented. The EU Markets in Crypto‑Assets regime applied core authorization from December 2024 and the transitional period for existing providers ended on July 1, 2026. The U.S. GENIUS Act was signed into law in July 2025 and entered a rulemaking phase by mid‑2026. FATF guidance and regional initiatives in Singapore and Hong Kong are also shaping verification requirements.

Third, some crypto‑native counterparties do not fit traditional registries. Decentralized autonomous organizations and multi‑signature treasuries often lack registered addresses or formal filings. Industry data cited in the report shows 25–35% of crypto users abandon onboarding when asked for an identity document and a selfie, a drop‑off that intensifies for multi‑party business verification.

The report says artificial intelligence is changing both fraud and detection. Generative models enable rapid creation of synthetic business identities, including fabricated incorporation documents, websites and deepfaked teams. The report notes synthetic business identity fraud is now the most common form of identity fraud globally.

At the same time, platforms use AI for behavioral and network detection. Techniques described in the report include submission forensics to detect metadata or editing traces, real‑time cross‑referencing of registries and websites, and link analysis that maps shared attributes such as addresses, tax IDs and device fingerprints to reveal clusters of linked accounts.

Persona documents operational approaches that platforms use to reduce friction and manage risk. Examples include gating KYB checks at fiat funding points before value is tokenized, embedding automated triggers for ongoing screening at the transaction level, and storing audit‑ready data structures from the start.

On reusable verification, the report distinguishes portable data from portable decisions. Sharing verified documents with user consent is already feasible, but accepting another firm’s verification outcome would require industry agreement on shared risk thresholds. The report identifies the lack of a common risk standard as the main barrier to portable verification decisions.

The report presents these findings without prescribing specific products. It records that some platforms adopt continuous verification and signal‑based detection while routing higher‑risk cases to manual review. Those approaches are described as practices used by platforms that manage changing requirements and fraud patterns.