AI Tenant Screening Compliance: What Colorado Property Managers Must Know About SB 24-205

If you're a Colorado property manager using AI-powered tenant screening — and nearly all modern screening services use AI — you're squarely in SB 24-205's crosshairs. Housing is explicitly listed as a "consequential decision" domain. AI-driven tenant screening is perhaps the clearest example of consequential AI: it directly determines who gets housing and who doesn't. Every screening decision is a potential compliance event. The overlap with Fair Housing Act obligations makes this doubly sensitive. AI that produces racially disparate screening outcomes isn't just an SB 24-205 problem — it's a federal fair housing problem.

Related: SB 24-205 compliance guide · bias audit guide · real examples of AI discrimination

**Tenant screening AI:** - Credit-based scoring algorithms - Criminal background check AI (risk scoring, not just record retrieval) - Eviction history analysis - Income verification and rent-to-income ratio tools - "Tenant quality" composite scoring **Other property management AI:** - Dynamic pricing/yield management for rental rates - Automated lease renewal recommendations - Maintenance request prioritization - Marketing AI (who sees your listings) Every one of these systems that affects a housing decision requires SB 24-205 compliance. The screening AI is the highest-risk category.

Tenant screening AI combines multiple data sources that individually correlate with race, ethnicity, and socioeconomic status: **Credit scores:** Well-documented racial disparities in credit scoring. Black and Hispanic applicants have lower average credit scores due to structural factors, not individual creditworthiness. **Criminal background:** Racial disparities in the criminal justice system flow directly into screening AI. Using criminal history as an input amplifies existing systemic bias. **Eviction history:** Eviction rates are significantly higher for minority communities, often due to predatory landlord practices and housing instability — not tenant quality. **Income requirements:** Rigid rent-to-income ratios disproportionately affect communities with lower average wages due to occupational segregation. When AI combines these inputs into a single "tenant quality score," the biases compound. A system using credit + criminal + eviction + income can produce dramatically different approval rates across racial groups — even without using race as an input. This is exactly the kind of proxy discrimination SB 24-205 was designed to address.

**1. Bias audit your screening AI:** Test approval/denial rates across all protected classes. Apply the four-fifths rule. If any demographic group is approved at less than 80% the rate of the highest-approved group, adverse impact is indicated. **2. Provide consumer disclosures:** Before running AI screening: Tell applicants that AI will be used in the screening process. After an adverse decision: Tell the denied applicant what factors contributed, that AI was involved, and how to contest the decision. **3. Offer a contest process:** Applicants denied by AI screening must have a way to request human review. This can't be a rubber stamp — the human must have authority to override the AI. **4. Publish your AI governance policy:** Your website must include a statement describing how you use AI in tenant decisions and how you manage associated risks. **5. Maintain records:** Keep all screening decisions, bias audit results, consumer disclosures, and contest outcomes for 3+ years. **6. Report discrimination:** If your bias audit reveals algorithmic discrimination, notify the Colorado AG within 90 days.

SB 24-205 compliance doesn't replace Fair Housing Act obligations — it's additive. **Federal Fair Housing Act (FHA):** - Prohibits discrimination in housing based on race, color, national origin, religion, sex, familial status, and disability - Disparate impact theory applies — facially neutral policies that produce discriminatory effects violate the FHA - HUD guidance specifically addresses algorithmic discrimination in tenant screening **Colorado Fair Housing Act:** - Adds additional protected classes: ancestry, creed, sexual orientation, gender identity, marital status, lawful source of income - Colorado Civil Rights Division enforcement **SB 24-205:** - Adds AI-specific obligations: bias audits, consumer notices, AG notification, evidence documentation A property manager using AI screening must comply with all three frameworks simultaneously. The good news: CO-AIMS's bias auditing and evidence documentation supports compliance across all three. Document your bias audits and remediation efforts, and you've built a defense for federal, state civil rights, and SB 24-205 claims.

Most property managers don't build their own AI — they use screening services like TransUnion SmartMove, RentPrep, AppFolio, or Buildium. You're still the "deployer" under SB 24-205. **Your obligations as deployer:** - Request bias testing data from your screening vendor (they have "developer" obligations under SB 24-205) - Conduct your own bias audits on the screening outcomes for your specific applicant pool - Provide consumer disclosures to applicants (the screening vendor doesn't do this for you) - Maintain your own evidence trail **Ask your screening vendor:** 1. Have you conducted bias audits on your AI models? 2. Can you provide disparate impact data for Colorado applicants? 3. Do your models use criminal history, eviction history, or credit scores? (All proxy-risk inputs) 4. Will you provide documentation for our SB 24-205 compliance? If your vendor can't answer these questions, you may need to supplement with independent bias testing. CO-AIMS can help you audit the outcomes of third-party screening AI.