The AI Tenant Risk Score Explained: What It Is, How It Works, and Why It's Better Than a Credit Score Alone

Why a Credit Score Is Not a Risk Score

Credit scores were designed by creditors for creditors. FICO was built to predict whether someone will repay a loan — not whether they'll be a good tenant. A 680 credit score could represent someone who had one medical collection four years ago that's been fully paid, with an otherwise spotless payment record across seven accounts. Or it could represent someone in a rolling cycle of missed credit card payments who is always two months behind. The number alone doesn't tell you which. And for a landlord, the distinction matters enormously. Medical collections have almost no predictive value for tenancy. Rent-related payment behavior has very high predictive value. An AI risk score reads the full credit report — not just the three-digit summary — and interprets the data through the lens of tenant behavior specifically.

The Four Data Sources in an AI Tenant Risk Score

1. Credit Report

Not just the score — the full tradeline analysis. This includes payment history across every open and closed account, length of credit history, types of accounts (revolving vs. installment vs. open), recent hard inquiries, and the specific nature of any collection accounts. AI distinguishes between a medical collection (low predictive weight for tenancy), a utility collection (moderate weight), and a prior landlord judgment (very high weight). It looks at whether derogatory marks are recent or historical, isolated or patterned. A single 30-day late payment from 2021 on a single account is categorically different from three 60-day lates in the past 18 months across multiple accounts.

2. Criminal Background Check

A nationwide criminal database search returns offense type, severity, jurisdiction, date, and disposition. AI weighs these contextually. A 2009 misdemeanor possession charge with no subsequent offenses is a different risk profile than a 2023 felony fraud conviction. The AI also applies jurisdictional rules: many states and cities have 'fair chance' or 'ban the box' ordinances that restrict what criminal history can be considered, and in some cases, what can be asked. VerticalRent's AI is aware of these restrictions and flags when certain offense types cannot legally be used in the applicant's jurisdiction.

3. Eviction History

Prior eviction records are arguably the single most predictive factor for future eviction risk. Court filings, judgments, and dismissals all tell a story. AI distinguishes between an eviction filing that was dismissed (which may represent a landlord error or resolved dispute), a default judgment (tenant didn't show up to court), and a judgment with an accompanying money award (serious non-payment event). Timing matters enormously: an eviction from 2019 that was followed by four years of clean rental history has much lower predictive weight than one from 2024. The AI scores these factors explicitly.

4. Rental History / SSN Trace

Address history verified through SSN trace gives you a picture of an applicant's residential stability. How many addresses in how many years? How long were they at each residence? Gaps between addresses (possible couch-surfing or unstable housing) are flagged. An applicant with six addresses in three years is a different risk profile than one with two addresses in seven years. This data also cross-references rental payment references from prior landlords where available. Stability in housing history is one of the stronger positive indicators in tenant screening.

How the 0–100 Score Is Calculated

The AI synthesizes all four data sources simultaneously. Eviction history and payment behavior patterns carry the heaviest weight. Criminal background is weighted by offense type, severity, and recency. Rental history stability adds confidence in either direction. Credit tradeline analysis provides the nuanced financial picture. Income-to-rent ratio from the application is layered in as a final check. All of this produces a score on a 0–100 scale, with a written explanation that tells the landlord exactly what drove the score — what was positive, what was negative, and what the key decision factors are.

What Landlords Can and Can't Do With the Risk Score

The AI Risk Score is a decision-support tool, not a decision-maker. Under the Fair Housing Act and FCRA, landlords must apply consistent, objective criteria to all applicants and cannot use protected class characteristics in any screening decision. The risk score helps you apply consistent criteria — the same scoring logic runs on every applicant, eliminating subjective variation. However, you must still exercise judgment. You cannot use criminal history in jurisdictions with fair chance ordinances. If you deny an applicant based on screening results, you must issue an adverse action notice informing them of their right to dispute the report. VerticalRent automatically generates adverse action notices — one less thing to remember in a legally sensitive moment.

AI Risk Scores vs. Traditional Screening Criteria

• Traditional: minimum credit score (e.g., 620+), income 3x rent, no evictions — pass/fail with no nuance
• AI: synthesizes all factors into a score plus explanation, surfaces nuance, helps landlord make an informed judgment rather than applying blunt cutoffs
• AI doesn't discriminate: it applies the same criteria to every applicant consistently, reducing Fair Housing risk
• AI explains its reasoning: you can document your decision-making process if a tenant challenges a denial
• AI catches what rules miss: a 660 credit score from a recent medical bankruptcy looks very different from a 660 from persistent non-payment — rules can't distinguish them

How to Use the AI Risk Score in Your Screening Process

1. 1Set your minimum score threshold before you start screening — for example, accept applicants scoring 60 or above
2. 2Read the AI explanation, not just the number — the reasoning tells you what drove the score
3. 3Review any flagged items manually — the AI will call out things that deserve a closer look
4. 4Apply the same standard to every applicant (Fair Housing compliance requires consistent criteria)
5. 5Issue an adverse action notice if denying — VerticalRent generates this automatically based on the screening result