FFIEC Authentication GuidanceAccess ControlsMultifactor AuthenticationChurch Extension FundsRisk Assessment

FFIEC Authentication Guidance for Church Funds

By • • • 16 min read
FFIEC Authentication Guidance for Church Funds

The exam letter arrives in March. It asks for the fund's authentication policy, the risk assessment behind it, and evidence that the controls were tested. On the same desk sit the investor note register, a stack of loan payment schedules, and a cash position someone rebuilt by hand on Friday afternoon.

Most church extension funds can answer the first question from memory. The second question is where the room gets quiet. FFIEC authentication guidance has never required a particular product, and that is precisely why it's hard to prepare for. It asks management to prove a chain of reasoning: who can reach the records, what harm follows if that access is misused, which control answers the risk, and how the fund knows the control still works.

What follows is a working method for the finance leader who has to assemble that file, brief a board, and get back to the loan pipeline.

What the 2021 Guidance Asks a Fund to Prove

Start with the document itself. The current standard is Authentication and Access to Financial Institution Services and Systems, issued in 2021. It replaced two earlier publications: Authentication in an Internet Banking Environment from 2005 and its 2011 supplement. The replacement matters because the current text reaches well past a member's login screen. It addresses identification, authentication, and access for customers, employees, and third parties using digital banking services and internal information systems.

It is also risk-based rather than prescriptive. Management evaluates the fund's services, users, transactions, systems, and threat environment, then selects controls appropriate to the risks identified. Nobody hands you an approved product list. You get a standard of reasoning and a demand for evidence.

That changes the shape of the exam file. A single house rule ("everyone uses a password and a code") is not a risk assessment. The guidance expects the fund to distinguish between an investor viewing account information, an employee posting a loan payment, a treasury user initiating an ACH file, and an administrator changing permissions. Each may warrant different authentication strength and a different approval path.

Expectation Applies to Evidence to keep
Inventory every access population Investors, employees, and third parties Current user list by role and system
Classify activity by inherent risk Services, systems, and workflows Written ratings for ACH, ledger posting, notes, loans, and admin changes
Use MFA or equivalent strength where single-factor risk is too high Groups flagged by the assessment Documented factors from different categories, not a second password
Detect and respond to suspicious activity Digital access and transactions Alert rules, review logs, and response notes
Reassess controls on a set cycle Management and the board Periodic review, residual-risk approval, and test results
Educate account holders Investors and congregations Notices on phishing and changes to payment instructions

One row at a time, with a fund in view

Take the table row by row and the abstraction dissolves.

An investor checking a note balance and downloading a statement is a read-mostly exposure. Account data, tax data, contact details. The control set is a unique identity, a session timeout, and MFA where the fund's own rating says the data warrants it.

A treasury user building an ACH file can move cash out the door in a single batch. That is a different animal, and the file should say so.

An employee posting a loan payment can misstate balances and, by extension, the investor reports built from them. An administrator changing permissions can hand broad access to anyone. Both deserve controls proportionate to the consequence, not to the job title.

The test that matters most: name the owner of each row and the file where the evidence lives. Any row without both is the work plan.

A fund with $40 million in notes and weekly redemption requests carries a different cash-movement exposure than a fund whose board only downloads a portfolio report once a quarter. Same guidance, different ratings, different evidence. If you want a structured reading of the examination handbook sections behind this, this guide to the FFIEC IT handbook walks through the expectations line by line.

The evidence set is not exotic. A current user inventory, written risk ratings, the authentication strength chosen for each group, notes from anomaly reviews, documented approval of residual risk, a review cycle with dates, and plain-language notices to investors about phishing and changes to payment instructions.

Does the Guidance Apply to a Church Extension Fund

Church extension funds are not FDIC-insured depositories. They operate under state securities rules and raise money through note programs. So the first question a finance leader asks is fair: does FFIEC guidance bind us directly, or is it just the benchmark everyone else uses?

The honest answer depends on the fund's charter, its regulator, and its contracts. The guidance is written for institutions supervised by the FFIEC member agencies. A fund outside that perimeter is not automatically subject to it. But being outside the perimeter is not the same as being unaffected.

Four steps settle the question for your board.

  1. Ask the state securities regulator. Confirm whether authentication or access-control standards appear in examination requests, questionnaires, or the examiner's manual. Ask for it in writing if you can.
  2. List every bank and processor you depend on. Deposit accounts, ACH origination, card and check services. Those partners are generally bound by FFIEC expectations and pass them down through contracts, annual certifications, and due diligence questionnaires.
  3. Ask the external auditor which framework is used to test access controls and exactly what evidence gets requested. Auditors usually arrive with a list already.
  4. Record the conclusion in board minutes, with the reasoning and a next review date.

The ACH example is the one that surprises people. A fund that originates payment files through a partner bank may find the bank requires MFA for every user who can create a file, even though the fund itself is not a federal institution. The obligation travels through the contract, not the charter.

Treating the guidance as your benchmark while separately confirming what binds you gives a board a defensible position. It avoids overstating legal obligations and avoids the opposite error, assuming a small fund is invisible.

How the Guidance Changed from 2005 to 2021

Boards still cite the wrong document. A director who served on a finance committee in 2010 may remember "the FFIEC booklet" and assume nothing has moved. Three milestones fix the timeline.

The first line below sets the progression in order, then a short table shows which document to cite when.

A timeline graphic showing the evolution of FFIEC authentication guidance from 2005 through 2011 to 2021.

In October 2005, the FFIEC issued Authentication in an Internet Banking Environment, aimed mainly at customer-facing online banking. Passwords were the norm and the document reflected it. On June 29, 2011, a supplement raised the bar and gave institutions a compliance deadline of January 1, 2012, roughly six months to upgrade controls, according to the FFIEC's 2021 announcement of the replacement guidance.

That 2011 supplement carried the expectations that still sound familiar: annual risk assessments, layered security for consumer accounts, stronger layered controls proportionate to the greater risks of business accounts, and more active customer education. It also told institutions to detect and respond to suspicious or anomalous activity, and it stated that certain device-identification methods and challenge questions should no longer be treated as effective controls on their own.

Date Document Status and scope
October 2005 Authentication in an Internet Banking Environment Customer internet banking; superseded in 2021
June 29, 2011 Supplement to Authentication in an Internet Banking Environment Annual risk assessments, MFA offered to business customers, compliance deadline January 1, 2012; superseded in 2021
2021 Authentication and Access to Financial Institution Services and Systems Customers, employees and third parties; current benchmark

The 2021 document did not add a third supplement. It replaced both earlier documents, which is the detail boards miss. The scope widened from online banking customers to customers, employees, and third parties reaching digital services and information systems.

One idea from 2011 survived intact and applies directly to fund operations: higher transaction values and broader administrative rights justify stronger controls. A staff member who can move cash or rewrite a record sits closer to that standard than a member checking a note balance. Keep the old PDFs as history. For a current examination file, the 2021 guidance is the benchmark, supported by documented assessments, appropriate controls, anomaly response, and proof that safeguards are reviewed and improved.

Layered Security and What Counts as MFA

Think of a lockbox with two latches on different faces. A password is one kind of latch, something only the user knows. A second copy of the same secret is still the same kind of latch. That distinction is where most fund policies fall apart.

The FFIEC framework defines layered security as a combination of preventive, detective, and corrective controls designed to compensate for weaknesses in any single control, as set out in the 2021 authentication and access guidance. Prevention stops access before it happens. Detection notices access that should not have occurred. Correction fixes and recovers after an incident. A fund with strong prevention and no detection has one working layer.

The three factor categories

The 2021 guidance highlights the weakness of single-factor authentication and states that multi-factor authentication, or controls of equivalent strength, can more effectively mitigate authentication and access risks when risk assessments show that single-factor controls with layering are inadequate.

MFA requires at least two distinct factors, drawn from three categories:

  • Something the user knows. A password, a PIN, an answer.
  • Something the user possesses. A hardware token, a digital certificate on a device, an authenticator app generating codes.
  • Something the user is. A fingerprint or other biometric.

Two passwords, or a password plus a security question the user also chose, cross no category boundary. That is one latch twice.

An infographic explaining the concept of multi-factor authentication, layered security, and its preventive, detective, and corrective components.

The method also has to survive the attacks in play. Assess the chosen control against credential theft, malware, and man-in-the-middle attacks. A code typed from a text message can be relayed in real time by an attacker sitting between the user and the login page. For teams weighing the options, these multi-factor authentication best practices lay out the practical differences between app-based codes, hardware keys, and passkeys.

Who gets the stronger control

The institution still has to do the analysis. Identify all categories of users and customers needing access controls. Determine which groups warrant enhanced safeguards such as MFA. Apply controls proportionate to the assessed risk. Then evaluate periodically whether the control is still effective, which is the part most funds skip.

For high-risk customer transactions, related FFIEC assessment materials cite visual transaction signing as an example of a control designed to prevent manipulation of transaction details before the user authorizes them. The point is that the payment itself gets checked, not just the login.

Not every investor needs a hardware token on day one. Every investor group needs a documented reason for whatever control it has.

A defensible design combines MFA or an equivalent control with role-based access, transaction monitoring, anomaly response, periodic testing, and audit evidence. Passwords alone are not that design, and neither is a policy that says "MFA required" without naming which factor satisfies it. Our own note on security in layers covers how those controls stack in a fund's system design.

How to Rate Access Risk for CEF Workflows

FFIEC authentication guidance asks for a risk-based program covering customers, employees, and third parties, not a login-control checklist. The useful output is a document a controller and an IT director can finish in two working sessions.

A five-step infographic showing a guide on how to rate access risk for CEF workflows in business environments.

Here is the sequence, adapted to fund work.

  1. List every population that can reach the records. Investors, borrowers, loan staff, accounting staff, treasury, administrators, board members with system access, outside processors and vendors.
  2. List actions, not job titles. View an investor statement. Change a bank instruction. Post a loan payment. Approve a construction draw. Create an ACH file. Post to the general ledger. Release a scheduled job. Grant permissions. Titles hide access; actions expose it.
  3. Rate each action by what could go wrong, measured in dollars and in trust. A misdirected $250,000 construction draw, or a redeemed note paid to the wrong account, outranks a read-only portfolio report. Rate high, medium, or low and write the reason next to the rating.
  4. Mark who needs enhanced authentication and where a second person should approve the action before it takes effect.
  5. Compare the control you have today with that rating and write the gap down plainly. A shared login, an unlogged spreadsheet edit, a generic administrator account used by three people.
  6. Have management approve the residual risk against the fund's risk appetite and policies, with a date for the next review.
  7. Keep the ratings, exceptions, and test results together and reopen the file whenever a new portal, ACH service, vendor, or remote-work arrangement changes the threat.

The cause and effect is simple. Authentication strength has to rise when exposure rises. A system supporting ACH initiation, investor-record access, loan disbursement, or privileged administration presents materially greater consequences than read-only reporting, and the FFIEC authentication and access guidance frames the assessment around exactly that relationship.

A shortcut worth resisting: don't start with the control you already own and write a rating that justifies it. Start with the harm, then pick the control.

The guidance does not prescribe one universal technology. It expects management to select controls whose remaining risk is consistent with the institution's risk appetite and policies. That opens the door to adaptive controls: step-up MFA for sensitive actions, separate privileged accounts, device or session risk signals, and maker-checker approval for high-impact transactions.

If you need a scaffold for the document itself, this cyber risk assessment template gives you headings that map cleanly onto access risk, and our overview of access control best practices shows how the ratings translate into system roles. Documented risk ratings, approval of residual risk, periodic reviews, exception records, and testing results are the evidence a reviewer will ask to see.

Controls that Fit Each User and Transaction

Apply the risk assessment to the fund's actual work. The table compares common tasks, the harm misuse could cause, and a control pattern suited to each.

User or action Exposure if misused Minimum control pattern
Investor views a statement Account, tax, and contact data Unique identity, timeout, and MFA when the rating requires it
Employee posts a loan payment Misstated balances and investor reports Named user, MFA, and a role limited to posting
Treasury creates an ACH file Cash leaving the fund MFA, separated duties, and maker-checker approval
Administrator changes permissions Broad access granted to anyone Separate admin account, MFA, and logged approval
Vendor or processor connects Wide reach into systems and files Expiring access, MFA, contract limits, and monitoring
Payee or bank instruction changes Redirected draws or note redemptions Step-up authentication and review of payment details

The highest practical risk may be a change to payment instructions, not creation of an ACH file. A stolen credential can be used to edit the destination for a construction draw or note redemption. The transaction may then look valid in the audit trail because it appears under a staff member's account. Require a step-up check and a human review of the payee details for every such change, regardless of who requests it.

The 2021 FFIEC Authentication and Access to Financial Institution Services and Systems guidance distinguishes MFA from repeated use of the same kind of secret: the authentication factors must be distinct and drawn from different categories. Its examples include tokens and digital certificates. The guidance also emphasizes stronger methods for employees accessing high-risk systems and for third-party access to institutional networks, systems, and applications.

That distinction matters when evaluating a platform. A password plus a second password is not meaningfully equivalent to a password paired with a hardware token, certificate, authenticator application, or biometric factor. A shared mailbox login is not a substitute either, nor is an administrator account protected only by a strong password.

Controls should fit the task. Read-only statement and tax-data access calls for a unique identity and a timeout, while treasury staff creating payment files need separation between the person who builds a file and the person who releases it. Vendor access should expire on a date recorded in the contract. For each row, the fund should be able to demonstrate the control on request, including who had access and how sensitive actions were reviewed.

What the Board Should Approve Next

Bring the file to the audit or finance committee and ask for a resolution, not a discussion. The motion has four parts.

Management presents the user inventory, the risk ratings for statement viewing, loan posting, ACH file creation, ledger posting, bank-instruction changes, and permission changes, and the list of groups receiving MFA or an equivalent control. Management also presents a signed acceptance of any residual risk that remains after those controls, with the reasoning behind it. The board sets the review cycle at least annually, and sooner whenever a payment channel, portal, or vendor is added. The board requires separate privileged accounts and a second approver for every cash movement and every payee change.

Then add a retirement list to the next policy review. These should not survive it.

  • The 2005 booklet as a current benchmark. It's history, not a standard.
  • Two passwords described as MFA. Same latch, twice.
  • Challenge questions or device identification used alone. The 2011 supplement already stopped treating these as sufficient.
  • Shared spreadsheet logins or a generic administrator account.
  • Unlogged changes to a church's deposit account or payment instructions.

Investor education can be modest. A short notice in the next statement cycle about unexpected requests to change payment instructions covers the expectation without a campaign. On the platform side, MFA is configurable for privileged accounts, TOTP authenticator apps are supported, and runtime enforcement is not currently active, which means your written policy, not the software default, has to carry the enforcement. CEFCore's controls are designed to meet or exceed SOC 2 Type II Trust Service Criteria, NIST CSF 2.0, and IRS Publication 1075 requirements, and customers may request a summary of the most recent security assessment results by contacting security@cefcore.com.

If the fund is replacing disconnected spreadsheets, judge any candidate system by one question. Does it keep named users, role limits, authentication events, and approvals in a single record you can hand an examiner? A control you can't evidence is a control you can't defend, whatever the policy says.


CEFCore was built for church extension funds that need named users, role-based access, and maker-checker approvals recorded in one auditable place, alongside the loan, note, and general ledger data those controls protect. If your board is ready to move from a spreadsheet-based access trail to a defensible record, visit CEFCore to see how the platform handles investor notes, ACH operations, and approval workflows for funds like yours.

CEF

CEF Core Editorial Team

Written and reviewed by CEF Core's treasury, fund-accounting, and compliance team — the people who build the financial management platform purpose-built for Church Extension Funds. Learn more about CEF Core.