The Coordination Tax

A mid-size lease-to-own lender in the industrial corridor didn't lose to a competitor. It didn't lose to a recession or a regulatory action. It lost to its own decisioning lag — the growing distance between what its systems could see and what the market was already doing around it. The Coordination Tax had arrived in consumer finance. Most lenders don't know they're paying it yet.

Part I

The Case of Meridian Lending

Meridian Lending — a composite drawn from patterns observed across multiple mid-market consumer finance operations — ran a clean shop by traditional measures. Delinquency rates within tolerance. Compliance calendar current. A servicing platform built in the mid-2010s that had processed hundreds of thousands of accounts without a major incident.

What Meridian could not see was velocity. The buy-now-pay-later providers entering its market were not just offering different terms. They were operating on a fundamentally different information cycle. Where Meridian's underwriting process took hours and touched legacy infrastructure at seven points, its emerging competitors were making decisions in seconds, feeding those decisions back into model training within 24 hours, and repricing risk continuously.

Meridian's models were not wrong. They were slow — and in an AI-native competitive environment, slow is a different kind of wrong. The decisioning gap was not a failure of intelligence. It was a failure of coordination between the intelligence the organization had and the infrastructure required to act on it in time.

By the time Meridian's leadership recognized the pattern, three things had happened simultaneously: their best-performing customer segment had shifted toward competitors offering instant decisioning; their delinquency models had begun producing false signals because the underlying consumer behavior profiles had changed faster than the training data could follow; and a state regulator had begun an inquiry into whether their AI-assisted underwriting process met the new adverse action explanation requirements. None of these were catastrophic individually. Together, they were a coordination failure — and the tax came due all at once.

Part II

The Signal Behind the Story

Meridian is a pattern, not an anomaly. The conditions that produced its coordination failure are structural — and they are present across consumer lending at scale.

The 88% adoption figure from the 2026 HAI AI Index represents organizations that have introduced AI tools into their workflows in some meaningful way. The 8% figure represents those that have deployed AI with genuine operational authority — systems that make or materially influence real decisions without a human in the loop for each transaction.

That gap — eighty points — is the Coordination Tax breeding ground. The organizations in the middle have AI capability without AI infrastructure. They have models without governance rails. They have adoption without accountability architecture. And in consumer finance, that middle ground carries specific legal exposure that does not exist in less regulated industries.

Part III

Why Consumer Finance Is Uniquely Exposed

The Coordination Tax exists across industries. But consumer finance carries a version of it that other sectors do not — because the regulatory accountability layer sits directly on top of the AI decisioning layer with no buffer between them.

The adverse action problem

When a model denies a credit application, existing fair lending law requires an explanation. Not a model explanation in the technical sense — a consumer-facing explanation, in plain language, that reflects the actual factors that drove the decision. The gap between what a model can technically output and what satisfies an adverse action notice under ECOA and FCRA is where many lenders are currently exposed without knowing it.

Seven states have introduced or passed AI disclosure legislation that extends this requirement further — mandating that consumers be informed when AI materially influenced a credit decision, and in some cases requiring that a human review be available on request. The federal framework is following, not leading, this movement.

The model drift problem

Consumer behavior has shifted faster in the past three years than in the prior decade. BNPL normalized multi-lender exposure for thin-file borrowers. Inflation cycles compressed the time between income disruption and default. Gig economy income patterns broke the payroll-to-risk correlation that underwriting models were trained on. The lenders whose models were last retrained in 2022 or 2023 are not just working with stale data — they are working with a map of a geography that has materially changed.

The infrastructure gap problem

The Coordination Tax compounds for organizations running decisioning infrastructure built before the current AI cycle. The systems were designed for a world where models were updated quarterly and human review was the default for edge cases. In the current environment, edge cases are the majority of interesting decisions — and the infrastructure was never built to route them correctly.

Retrofitting accountability onto infrastructure that was not designed for it is not a technology problem. It is an architecture problem. And architecture problems are expensive to solve after the fact.

Part IV

What Readiness Actually Looks Like

The organizations that will not pay the Coordination Tax — or will pay it minimally — share a set of structural characteristics. None of them are about having the most sophisticated AI. All of them are about having the right infrastructure underneath it.

• 1
  Decisioning infrastructure built for accountability, not just throughput Every decision the system makes is logged with the factors that drove it, in a format that can produce a compliant adverse action notice without post-processing. This is not a reporting feature. It is an architectural requirement that must be built in from the beginning.
• 2
  Model governance on a continuous cycle, not a quarterly calendar The organizations absorbing AI disruption are retraining on shorter cycles, with drift detection that surfaces behavioral shifts before they become model failures. The governance calendar follows the data, not the fiscal quarter.
• 3
  Multi-bureau and alternative data integration as a baseline BNPL exposure, rent payment history, and gig income signals are now material underwriting inputs for a significant share of the applicant pool. Organizations treating these as optional enrichment rather than baseline data are operating with a structural blind spot.
• 4
  Regulatory monitoring that runs ahead of the compliance calendar The seven-state AI disclosure movement is a leading indicator, not an isolated development. The organizations building disclosure capability now — before it is legally required in their operating states — will not be in retrofit mode when the mandate arrives.
• 5
  Intelligence ownership rather than vendor dependency The lenders most exposed to the Coordination Tax are those whose AI capability lives entirely in third-party systems they do not control, cannot explain, and cannot modify. Sovereignty over the decisioning layer is not an ideological position — it is a regulatory and competitive necessity.

Conclusion

The Operators Who Understand This Will Inherit the Market

Meridian's story does not end in failure. The pattern ends in failure when organizations encounter it without a framework for recognizing what is happening. With the framework, it ends in a different kind of decision: build the accountability infrastructure now, before the regulatory mandate arrives and before the competitive gap becomes unrecoverable.

The Coordination Tax is a payment schedule, not a sentence. The lenders who read the signals early — who understand that AI readiness is not about tool adoption but about governance architecture — are building the infrastructure that will be the moat in the next competitive cycle.

The AI-native lending era is not coming. It is here. The question for every consumer finance operator is whether the infrastructure underneath their decisioning is ready to carry the accountability that comes with it.