What Compliance Work Actually Stays With You in a Banking-as-a-Service Program · Midcore Operations
Midcore Operations
Payments & Fintech Advisory

What Compliance Work Actually Stays With You in a Banking-as-a-Service Program

A sponsor bank's own guidance treats complaint handling, marketing review, and KYC collection as negotiated terms, not assumptions left unsaid.

Published
9 November 2026
Reading time
10 min read
Author
Payments & Fintech Advisory practice
Topic
Payments & Fintech Advisory
Key takeaways
  • The 2023 interagency guidance on third-party relationships treats customer complaint handling as a specific contract term. The bank and the third party have to explicitly allocate who responds, and if it is the third party, the contract has to require timely response plus trend reporting back to the bank.
  • The same guidance's planning stage directs a bank to assess a prospective third party's potential for consumer harm and its interaction with customers before the relationship even starts. That is exactly why a program's marketing materials and customer-facing disclosures get reviewed as part of the bank's own diligence, not as an afterthought the fintech handles alone.
  • Federal Customer Identification Program rules set the minimum identifying information a bank must collect before opening an account. In a banking-as-a-service program, the fintech's onboarding flow usually performs that first-line collection on the bank's behalf, which is why KYC data collection shows up as fintech-side work even though the bank's CIP legally owns the result.
  • None of this is static once the program launches. The guidance's ongoing monitoring stage specifically tracks the volume, nature, and trends of customer complaints and the adequacy of whichever party is handling them, so a complaint-handling gap discovered at launch does not stay hidden. It surfaces in the bank's recurring review.

What compliance work actually stays with you in a banking-as-a-service program comes down to a short list. The federal guidance governing your sponsor bank already answers it, item by item, whether or not anyone writes it down before launch. Customer complaints, marketing review, and first-line KYC collection are not left to assumption in a well-run program. They are specific allocations a bank’s own risk management process is built to make explicit.

Customer complaints: a contract term, not an assumption

The 2023 interagency guidance on third-party relationships, issued jointly by the Federal Reserve, the FDIC, and the OCC, addresses customer complaints directly in its contract negotiation guidance. It states that effective contracts typically specify whether the banking organization or the third party is responsible for responding to customer complaints and inquiries. If that responsibility sits with the third party, the contract needs real provisions behind it. It has to require the third party to receive and respond to complaints in a timely manner, and to provide the bank with sufficient, timely, and usable information to analyze complaint activity and trends.

This is precisely the mechanism behind why a fintech operating a BaaS program typically ends up handling first-line complaints. It is not that the bank has quietly outsourced its obligations. It is that the contract underneath the program has to make an explicit choice. In most programs, the fintech is positioned as the customer’s actual point of contact, which makes it the practical choice for who answers first. The obligation that stays with the fintech is not just answering the phone. It is maintaining the kind of complaint and inquiry records the bank actually needs to analyze trends on its own side.

Why “potential for consumer harm” puts marketing and disclosures on the radar

A second part of the same guidance explains something else. It explains why a bank’s diligence reaches into a fintech’s marketing materials and customer-facing disclosures before a program ever launches. In its planning guidance, the document directs a bank to assess a potential third party’s impact on customers specifically. That assessment covers the third party’s interaction with customers, its potential for consumer harm, and its handling of complaints and inquiries. That assessment happens during planning, before the bank has even selected a third party, not as a later compliance review bolted on after launch.

A fintech’s marketing claims, its account opening disclosures, and the language it uses to describe fees and features all sit inside what “potential for consumer harm” is actually asking a bank to evaluate. A program that has not prepared clean, accurate, substantiated marketing and disclosure language before the bank’s diligence begins is not just risking a slow approval. It is handing the bank evidence against the exact factor its own guidance tells it to weigh most carefully.

The identity data collection split

A third piece of compliance work stays with the fintech in practice, even though the underlying legal obligation sits with the bank. That piece is the first-line collection of identifying information from new customers. Federal Customer Identification Program rules, codified at 31 CFR 1020.220, require a bank to obtain, at minimum, a customer’s name, address, an identification number, and, for an individual, a date of birth, before opening an account. That obligation belongs to the bank’s own anti-money laundering compliance program.

In a banking-as-a-service program, the actual screens a customer sees belong to the fintech’s own onboarding flow, not the bank’s. The same is true of the actual form fields collecting that name, address, and identification number.

This split matters for a reason most fintechs underestimate until an application gets rejected. The bank’s CIP rule does not include a general provision letting a bank simply delegate its verification obligation to an unregulated technology company and treat the obligation as satisfied. The fintech’s onboarding flow is doing first-line data collection, not standing in the bank’s shoes under any formal reliance exemption. That means the quality of the data the fintech’s flow collects is the actual input the bank’s own verification process has to work with. A sloppy collection flow produces a verification failure the bank has every right to reject back to the fintech. The bank’s CIP program still owns the legal result: verifying that the information collected is accurate and sufficient. The fintech is doing the first-line work of actually gathering it correctly, completely, and in the format the bank’s verification process needs. A fintech’s onboarding flow that collects this information sloppily, missing a required field or allowing an obviously malformed entry through, is not creating a problem the bank quietly absorbs. It is creating rework and friction that lands back on the fintech the moment the bank’s verification process rejects the application.

Why none of this stays settled after launch

The same interagency guidance treats this allocation as something a bank has to keep checking, not a decision it makes once and forgets. Its ongoing monitoring guidance specifically calls out the volume, nature, and trends of customer inquiries and complaints as something a bank should track continuously. That includes the adequacy of the third party’s responses, if the third party is responsible for handling them. A fintech that handled complaints well at launch, with low volume and a small customer base, does not get to assume that performance holds steady as the program scales.

This is exactly the mechanism by which a compliance gap that seemed manageable at a small scale becomes visible to the bank later. A rising complaint volume, a pattern of slow responses, or trends suggesting a specific product feature is generating repeat complaints are all things the bank’s ongoing monitoring is specifically built to catch. A fintech that is not tracking this data as closely as its sponsor bank is, is the one most likely to be surprised when the bank’s own review flags it first.

The obligation that outlives the program itself

The same guidance also addresses what happens when a BaaS relationship ends, whether that is a planned transition to a new partner or an unplanned termination. Its termination guidance directs a bank to plan for managing risks associated with data retention and destruction. The same planning covers information system connections and access control, and other control concerns that require additional risk management after the relationship actually ends.

This is a compliance obligation most fintechs do not think about until the moment it becomes urgent. A program that switches BaaS providers, or winds down entirely, does not get to treat customer records, KYC documentation, and complaint history as the old partner’s problem to retain. The underlying data retention requirements do not pause just because the commercial relationship between the fintech and its BaaS partner changed. That includes the five-year retention period Customer Identification Program rules themselves specify for identifying information. A fintech that has not planned for where this data lives, and who can still access and produce it after a transition, is the party left exposed. That exposure shows up when a regulator or the new partner’s own diligence asks for records the old arrangement never clearly assigned anyone to keep.

A concrete example: the complaint that landed on the wrong desk

Picture a fintech running a BaaS-powered deposit account product, six months into launch. A customer emails a complaint about an unexpected fee directly to the fintech’s support inbox, which the fintech’s contract with its sponsor bank designates as the correct first point of contact. The support team resolves the individual complaint within a day, refunds the fee, and closes the ticket.

What the support team does not do is log that complaint in a way the bank can see trending alongside similar complaints from other customers hitting the same fee under the same circumstances. Three months later, the bank’s own ongoing monitoring review asks for complaint volume and trend data, exactly the kind of information the interagency guidance’s contract negotiation provisions anticipate the bank will need. The fintech can show the individual complaint was resolved quickly. It cannot show whether this was an isolated case or part of a pattern, because the complaint-handling process solved each ticket without ever building the aggregate record the bank’s monitoring actually depends on. The individual customer was taken care of. The compliance obligation that was supposed to accompany that work was not.

What a fintech actually needs ready before this gets negotiated

A fintech preparing to launch or renegotiate a BaaS program should walk into that conversation already knowing which side of each of these allocations it expects to hold, rather than discovering the answer inside whatever the partner’s standard contract happens to say. That means a complaint-handling process built from day one to produce the volume and trend data a sponsor bank’s ongoing monitoring will eventually ask for, not just a process that resolves individual tickets quickly. It means marketing and disclosure language reviewed and substantiated before diligence begins, since the bank’s own planning stage is specifically built to probe exactly that. And it means an onboarding flow that collects the Customer Identification Program’s required fields completely and correctly the first time, since every rejected application is friction that the fintech, not the bank, ultimately absorbs.

Mapping these allocations accurately before a program launches, rather than discovering the gaps during a bank’s diligence review or its first ongoing monitoring cycle, is exactly the work behind Midcore’s banking-as-a-service consulting engagements. Where the same compliance-allocation logic applies to a card program layered on top of the banking relationship, card-as-a-service consulting is where that parallel mapping gets done. And for a platform adding disbursement or lending accounts on the same banking rails, lending-as-a-service consulting is where the compliance split gets worked through for that specific use case.

Frequently Asked Questions

Why does a fintech, not the sponsor bank, usually handle customer complaints in a BaaS program?

Because that allocation is a negotiated contract term, not a default. The interagency guidance on third-party relationships states that effective contracts specify whether the banking organization or the third party is responsible for responding to customer complaints and inquiries. When the fintech takes that responsibility, the contract has to require it to respond promptly and to provide the bank with usable information on complaint volume and trends. The bank still has to monitor that activity, even when it is not the one answering the phone.

Does the sponsor bank actually review a fintech’s marketing materials before launch?

Yes, as part of its own due diligence obligation. The interagency guidance’s planning stage directs a bank to assess a prospective third party’s impact on customers, including its interaction with customers, potential for consumer harm, and handling of complaints and inquiries, before entering the relationship. A fintech’s marketing claims and customer-facing disclosures are squarely inside what that assessment is checking.

Who is actually responsible for collecting KYC information from a new customer?

Legally, the sponsoring bank’s Customer Identification Program owns the result: confirming the customer’s name, address, identification number, and date of birth for an individual before the account opens, under 31 CFR 1020.220. Operationally, the fintech’s own onboarding flow is usually the one actually collecting that information from the customer and passing it through to the bank’s verification process. That is why this shows up as fintech-side work in practice, even though the CIP obligation itself sits with the bank.

Sources: Interagency Guidance on Third-Party Relationships: Risk Management and 31 CFR 1020.220, Customer Identification Program Requirements for Banks.

Payments & Fintech Advisory practiceMidcore Operations · 9 November 2026
All insights
Free consultation

Get your free Assess

Tell us what is happening. We reply within one business day with a US-based practice lead, not a salesperson.

  1. 01Within 1 business dayA US-based practice lead replies and books a 30 minute call.
  2. 02On the callWe map the problem, the volume, the partners, and the constraints.
  3. 03After the callYou get a written read of one to three pages, yours to keep.
  • Handled under NDA
  • Written, not a slide deck
  • No obligation to continue

No obligation. We will tell you if you do not need us.

Before you go

Take the written read with you

Every engagement opens with an Assess: a written read of what is happening, what it costs, and what to do about it. It is quoted on its own and you can stop after it.

+1 (786) 619-0152