Skills Testing in Financial Services Hiring: Building a Defensible Program in 2026

Pre-employment skills tests remain one of the most practical ways for banks, insurers, credit unions and wealth firms to screen for accuracy, numeracy and judgment before an offer is made. The legal ground under those tests is shifting in 2026. On June 9, 2026, the U.S. Department of Justice told the Equal Employment Opportunity Commission (EEOC) that its disparate-impact guidelines are unconstitutional. On October 16, 2026, the EEOC sent the White House a final rule rescinding the interpretive portions of the Uniform Guidelines on Employee Selection Procedures (UGESP).

Title VII's statutory disparate-impact provision is unchanged, however. Private plaintiffs can still sue, and state laws are adding their own rules for AI-assisted assessments. For financial services employers, the practical conclusion matches what courts have said for five decades: a test that measures what the job actually requires, is applied consistently and is well documented is both a better hiring tool and a stronger legal position. This article explains how to build that kind of testing program for financial roles.

Why Financial Employers Test Skills Before Hiring

In financial services, small errors carry outsized costs. A misposted transaction, a reconciliation forced to balance, or a client explanation that misstates a fee can create losses, complaints and regulatory exposure. Resumes and interviews show what a candidate says they can do. A well-designed skills test shows what they can actually do.

Recent data shows the gap employers are trying to close. CFA Institute's Finance Skills Pulse Survey (April 2026) asked 1,350 finance professionals at manager level or above in the United States, United Kingdom and Canada what new entrants need on day one. Financial statement analysis led at 61%, followed by Excel at 47% and data analysis at 44%. Only 42% said new entrants are "very prepared" for the skills their roles require, and 53% named soft skills, such as communication and collaboration, as the biggest gap.

Those findings point to two testing priorities for financial roles: verify core technical skills directly, and measure transferable skills such as communication and judgment, not just what candidates know.

The Legal Framework Behind Employment Testing

The foundation. In Griggs v. Duke Power Co. (1971), the Supreme Court held that a facially neutral hiring requirement that disproportionately excludes a protected group violates Title VII unless it is job-related and consistent with business necessity. In Albemarle Paper Co. v. Moody (1975), the Court applied that rule to employment tests and rejected an aptitude test because the employer's validation study was materially defective. Congress wrote disparate-impact liability into Title VII itself in the Civil Rights Act of 1991, at 42 U.S.C. § 2000e-2(k).

The Uniform Guidelines. Adopted in 1978, UGESP gave employers a framework for testing. It treats a selection rate for any group below four-fifths (80%) of the highest group's rate as an indicator of adverse impact. When adverse impact appears, it recognizes three ways to show a test is job-related: criterion-related validity (test scores predict job performance), content validity (the test samples important job tasks), and construct validity (the test measures traits shown to matter for the job).

What changed in 2025 and 2026. Executive Order 14281, signed April 23, 2025, directed federal agencies to deprioritize disparate-impact enforcement. On June 9, 2026, the Justice Department issued an opinion concluding that the EEOC's disparate-impact guidelines are unconstitutional. The EEOC has since sent a final rule rescinding UGESP's interpretive portions for White House review, with an effective date projected for January 2027. A separate, earlier-stage proposal would remove UGESP's recordkeeping requirements, and current obligations generally stay in place until a final rule takes effect.

What has not changed. An executive order and an agency opinion cannot amend a statute. The disparate-impact provision in Title VII remains law, private plaintiffs can still bring claims, and state anti-discrimination laws and agencies operate independently of federal enforcement priorities. The Americans with Disabilities Act also still requires employers to give tests in a way that measures the skills being tested rather than a candidate's impairment, and to provide reasonable accommodations.

For employers, the safest reading is that federal enforcement pressure has eased, while litigation risk and state-level obligations have not. A job-related, documented test remains the best defense in either environment.

What the Case Law Teaches Employers

Courts have rarely faulted employers simply for testing. They fault tests that are arbitrary, harder than the job, or poorly documented. Four decisions show the pattern.

Case	What happened	Lesson for financial employers
Isabel v. City of Memphis (6th Cir. 2005)	A police promotion test used a cut score of 70, later lowered to 66, that was never validated. The court found it arbitrary and ordered promotions and back pay.	Set pass marks from job evidence, never by convenience or after seeing results.
EEOC v. Dial Corp. (8th Cir. 2006)	A strength test cut female hiring from 46% to 15%. The employer's validation evidence was undercut by testimony that the test was much harder than the actual job and that injuries had already been falling before the test was introduced.	A test must reflect the real job. An Excel test built around functions the role never uses invites the same challenge.
Williams v. Ford Motor Co. (6th Cir. 1999)	Ford's hourly selection test was upheld because a professional developer built it from a job analysis and produced data linking scores to job performance.	A documented job analysis and performance data are the strongest protection.
Ricci v. DeStefano (U.S. 2009)	A city discarded promotion exam results after seeing the racial outcomes. The Supreme Court held this was itself unlawful discrimination.	Decide how a test will be used before giving it, and don't throw out results afterward because of who passed.

Taken together, these cases describe the same discipline: define the job first, test what the job requires, set standards in advance, and keep the records that prove it.

Building a Defensible Testing Program for Financial Roles

1. Start with a job analysis. For each role, list the critical tasks, the tools used and the errors that would matter most. Interview strong performers and their managers. This document becomes the basis for every test decision and the first exhibit if a test is ever challenged.

2. Match the test to the tasks. Different financial roles call for different evidence.

Role	Critical tasks	Assessments that fit
Bank teller or branch associate	Cash handling, account queries, explaining fees	Numerical accuracy, attention to detail, customer service situational judgment
Bookkeeper or accounts payable/receivable	Recording entries, reconciliations, spotting duplicates	Bookkeeping knowledge, Excel simulation, data-entry accuracy
Financial or credit analyst	Reading statements, modeling, recommending decisions	Financial analysis, Excel and data simulations, numerical and logical reasoning
Client service or advisor support	Explaining products, handling complaints, documentation	Written business communication, customer service and situational judgment tests
Operations or compliance support	Following procedures, escalating exceptions	Attention to detail, procedure-based scenarios, time management

3. Customize rather than rely on generic tests. A test built for a generic "office" role rarely reflects a lender's systems or an insurer's workflows, and a mismatch between test and job is what undermined the employer's case in Dial. Assessment platforms such as eSkill let employers assemble financial-services assessments from Excel, bookkeeping, financial analysis, customer service and compliance-related content, and add realistic work simulations, so the test mirrors the role rather than a job title. Treat any compliance content as a measure of procedural awareness, not proof that a candidate knows the regulations.

4. Set cut scores before testing. Base pass marks on what successful employees in the role actually score, and document the reasoning. A cut score chosen after the fact, or adjusted to change who passes, was the core problem in Isabel.

5. Administer consistently. Give every candidate for the same role the same test, time limit and conditions. Use proctoring or integrity controls proportionate to the role, and handle accommodation requests through a defined process.

6. Monitor outcomes and keep records. Compare pass rates across groups at least annually, using the four-fifths comparison as an early warning rather than a legal conclusion. Keep the job analysis, test content, cut-score rationale and results even if federal recordkeeping rules are rescinded, because private plaintiffs and state agencies can still ask for them.

7. Check the test against performance. After six to twelve months, compare new hires' test scores with their performance reviews, error rates or productivity. If scores don't predict success, revise the test. This is good hiring practice and the core of criterion-related validity.

AI-Enabled Assessments and New State Rules

As federal enforcement recedes, state and local rules are filling part of the gap, especially where assessments use AI to score or rank candidates.

New York City. Local Law 144 requires an annual independent bias audit, candidate notice and a published summary of audit results before employers use an automated employment decision tool for NYC candidates.
Illinois. HB 3773 took effect January 1, 2026. It amends the Illinois Human Rights Act to bar AI use that has a discriminatory effect, prohibits using ZIP codes as a proxy for protected classes, and requires notice when AI is used in employment decisions. The state withdrew its proposed implementing rules on June 2, 2026, but the statute itself remains in force.
Colorado. The state replaced its original AI Act with SB 26-189, effective January 1, 2027, which imposes narrower notice and disclosure duties.

Before adopting any assessment with automated scoring, ask the vendor which features use AI, what bias testing has been done, and whether the tool supports the notices these jurisdictions require. Traditional skills tests scored against fixed answer keys raise fewer of these questions, but the same job-relatedness discipline applies to both.

Skills Tests Do Not Replace Licensing

For roles in the securities business, regulatory qualification is a separate requirement. Under FINRA Rule 1210, a representative's registration takes effect only after the person passes the Securities Industry Essentials (SIE) exam and the appropriate representative qualification exam, such as the Series 7. Passing the SIE alone does not qualify anyone for registration, and state requirements such as the Series 63 may apply on top of FINRA's.

Licensing and pre-hire testing answer different questions. A license confirms a person meets the regulatory standard for the role. A skills test shows whether they can reconcile an account, build a clean spreadsheet or explain a fee clearly on a busy day. Financial employers hiring for registered roles need both, and neither substitutes for the other.

A Checklist for Financial Services Employers

Employment tests done well remain a useful and lawful part of hiring, whatever happens to the federal guidelines. Before relying on a test for any financial role, confirm that you can answer yes to each of the following:

A written job analysis exists for the role and the test maps to its critical tasks.
The test difficulty reflects the actual job, not an idealized version of it.
Cut scores were set and documented before candidates were tested.
Every candidate for the role takes the same test under the same conditions.
A clear process exists for disability accommodation requests.
Pass rates are reviewed across groups at least once a year.
Any AI-scored features have been identified and checked against NYC, Illinois and Colorado requirements where they apply.
Test records are retained, and scores are periodically compared with on-the-job performance.
Registered roles still require the applicable FINRA and state qualifications.