Este artigo está disponível apenas em inglês no momento.
AI and Your Credit Score: How Algorithmic Lending Works and What Your Rights Are
AI systems now make or substantially influence most credit decisions. Understanding how they work, why they can go wrong, and what your legal rights are when you're denied credit by an algorithm.
Key Takeaways
Most credit decisions — mortgages, credit cards, personal loans, buy-now-pay-later — are now made or substantially influenced by AI models that analyse hundreds or thousands of data points about you.
These AI models can be wrong, biased, or based on data that doesn't reflect your actual creditworthiness — and unlike human decisions, they may be difficult for even the lender to explain.
In Australia, the Responsible Lending Obligations and Privacy Act create rights to receive reasons for credit denials and to access the data used to make them. In the EU/UK, GDPR Article 22 gives you the right to human review of automated credit decisions.
In the US, the Equal Credit Opportunity Act requires specific adverse action reasons for credit denials — 'our model declined you' is not sufficient, and the CFPB has actively enforced this for AI credit decisions.
Practical actions: request the specific reasons for any credit denial, check and correct your credit file, ask for a human review of automated decisions, and consider complaining to your regulator if you believe an AI credit decision was unfair or based on incorrect data.
"Apenas para fins informativos. Este artigo não constitui aconselhamento jurídico, regulatório, financeiro ou profissional. Consulte um especialista qualificado para orientação específica."
How AI credit decisions work
Modern AI credit scoring analyses far more data than traditional credit scores. Beyond payment history and existing debt, AI models may consider: the type of device you use to apply, the time of day you apply, how long you spent on the application, social media signals (in some markets), geolocation data, spending patterns from open banking data, and hundreds of other variables. These models find statistical correlations between these variables and loan default rates — and use those correlations to assess your creditworthiness.
The problem is that statistical correlations can encode discrimination without explicitly using protected characteristics. A model trained on historical data will reflect historical lending patterns — which in most markets have disadvantaged women, certain ethnic groups, and people in particular geographic areas. A postcode that correlates with a demographic group becomes a proxy variable for that demographic group, producing discriminatory outcomes through an apparently neutral algorithm.
Your rights by jurisdiction
In Australia: the National Consumer Credit Protection Act's responsible lending obligations require lenders to make inquiries before granting credit and to assess suitability. If you are denied credit, you can request the reasons. If the denial was based on credit report information, you can request access to your credit report, dispute inaccurate information, and ask the lender to reconsider. The Privacy Act gives you access to all personal data held about you including data used in credit assessments.
In the EU and UK: GDPR Article 22 applies to automated credit decisions. If your credit application was decided solely by automated processing (no genuine human review), you have the right to: request human intervention, express your view before the final decision, and contest decisions that have already been made. You can request the specific factors that contributed to the decision and their relative weight — the lender must provide a meaningful explanation, not a reference to an algorithm.
In the US: The Equal Credit Opportunity Act requires creditors to provide specific reasons for adverse actions. The CFPB has specifically stated that "our model declined you" or reference to a credit score without specific reasons does not satisfy ECOA requirements. You have the right to know the specific principal reasons for any credit denial — typically the top four factors that negatively affected your credit assessment.