Artificial Intelligence (AI) has been a concrete reality in the credit sector for over 20 years, with machine learning technology, for example. Today, with the expanded use of Generative AI, the sector is incorporating more and more innovations to generate value for companies, consumers and society.
At ANBC, we are closely following this evolution. The AI journey in credit is already showing tangible results: reduced operating costs, increased predictive accuracy, improved consumer experience and expanded access to responsible credit.
The application of AI in Brazilian bureaux has revolutionized processes such as:
- Immediate updating of the credit score, using alternative data and decisions in minutes.
- Real-time fraud detection, with machine learning analyzing behavioral patterns.
- Automation of statistical modeling, reducing costs by up to 66% and speeding up model development from weeks to days.
These advances are not just technical, they have direct implications for financial inclusion, credit sustainability and data governance.
AI applied with concrete results in Brazilian credit bureaus
In Brazil, bureaus are investing heavily in AI to transform the credit market with security, speed and responsibility.
In one of the companies, more than 70% of the new models developed in 2023 incorporated machine learning and deep learning, and its cloud infrastructure:
- 20x faster data processing
- Greater security for highly regulated data
- Agility in decision-making, with processes that used to take months being completed in days
In this case, AI is combined with human evaluation to ensure contextualized and effective decisions, such as detecting financial anomalies and consumption patterns.
Another Brazilian bank has implemented instant write-off, allowing the credit score be updated in seconds. This allows consumers to pay off their debt and have immediate access to new credit conditions, generating a structural change in the market.
In addition, the incorporation of data from social programs, telcos and utilities into predictive models has contributed to reduce defaults by up to 15%, without affecting approval levels.
Also on the national scene, the use of advanced analytical platforms has enabled a local bureau:
- Tripling the production of analytical models
- Reduce development time from 3-4 weeks to 2-5 days
- Increase the accuracy of the models by up to 8 points in the KS index, an indicator of predictive quality.
Another example of innovation in the sector is the use of models that analyze up to 24 months of account history, including actual payment amounts, to more accurately predict future consumer behavior.
These models offer:
- More than 300 premium predictive algorithms
- Analysis of payment behavior and balance magnitude
This granularity transforms credit decision-making, allowing for more personalized products, reducing defaults e responsible credit expansion.
AI governance is guided by ethical concepts within the bureaus, with extensive information security, monitored and transparent systems. Responsible innovation allows for increased efficiency and accuracy, fostering the inclusion and sustainability of credit in the country.
Regulation and responsibility: the role of ANBC
In addition to closely following national and international trends, ANBC has been active in debates on AI regulation, advocating a balanced approach that:
- Differentiate high-risk systems from traditional credit models.
- Avoid excessive interpretations of discrimination, in line with the LGPD.
- Promote decentralized governance and proportional penalties.
Thanks for reading! Access other content at ANBC website.
