Question 1

What is Attestant?

Accepted Answer

Attestant is an AI compliance platform for US banks and regulated financial institutions. It automates ML model validation against ECOA (Equal Credit Opportunity Act), SR 11-7 (Federal Reserve model risk guidance), EU AI Act, and Colorado AI Act (SB 24-205). It generates examiner-ready audit trails, fair lending analysis with disparate impact testing, AI-powered compliance reports, and deployment gating decisions — in minutes.

Question 2

What regulations does Attestant cover?

Accepted Answer

Attestant runs 54 automated checks across four regulations: ECOA/Reg B (14 checks including disparate impact testing and protected attribute isolation), SR 11-7 (15 checks including model inventory, validation framework, and champion/challenger testing), EU AI Act (13 checks including risk classification, transparency, and human oversight requirements), and Colorado AI Act SB 24-205 (12 checks including impact assessments, consumer disclosures, and appeal mechanisms). Checks are configurable via the Compliance Sandbox.

Question 3

How does Attestant test for fair lending compliance?

Accepted Answer

Attestant runs statistical disparate impact analysis using the 4/5ths rule (80% rule) across all 7 ECOA-protected classes: race, sex, age (40+), national origin, marital status, disability, and receipt of public assistance. It also verifies that protected attributes are mathematically isolated from model inputs and generates examiner-ready statistical exhibits formatted for ECOA/Reg B examination.

Question 4

How much does Attestant cost?

Accepted Answer

Attestant Starter is $199 per month with usage-based pricing: $0.25 per GB storage, $0.20 per GB egress, $10 per model validation, and $2 per pipeline run. All 54 compliance checks are included. Enterprise pricing is custom with additional features like vendor governance, SSO, and dedicated support.

Question 5

What is SR 11-7 and why does it matter for banks?

Accepted Answer

SR 11-7 is Federal Reserve supervisory guidance on model risk management for US banks. It requires banks to maintain a model inventory, conduct independent model validation, perform ongoing monitoring, and document model limitations and use restrictions. Non-compliance can result in Matters Requiring Attention (MRAs) or Matters Requiring Immediate Attention (MRIAs) from examiners. Attestant automates 15 SR 11-7 checks and generates validation packages formatted for examiner review.

Question 6

Does Attestant cover EU AI Act compliance?

Accepted Answer

Yes. Attestant includes 13 EU AI Act checks covering Annex III high-risk AI classification for credit scoring systems, Article 9 risk management, Article 10 data governance, Article 13 transparency documentation, and Article 14 human oversight requirements. EU AI Act enforcement for credit scoring AI systems begins August 2, 2026. The Compliance Sandbox lets you run EU AI Act checks independently of other regulations.

Question 7

Can Attestant integrate with existing ML pipelines?

Accepted Answer

Yes. Install the Python library with pip install attestant and call attestant.upload() to run compliance checks inside any ML workflow. A single function call runs fairness analysis, all 54 compliance checks, and returns a deployment gate decision. Results automatically sync to the dashboard. The library supports scikit-learn, XGBoost, LightGBM, and any model with a predict() method.

Question 8

What is the 4/5ths rule in fair lending?

Accepted Answer

The 4/5ths rule (also called the 80% rule) is a statistical test used by US bank regulators to detect lending discrimination. If the loan approval rate for a protected group is less than 80% of the approval rate for the most-favored group, the model may have a disparate impact violation under ECOA. Attestant automatically calculates disparate impact ratios for all protected classes and flags violations.