The Superintendency of Finance modernizes supervision with AI in its models

The Colombian Financial Superintendence (SFC) has launched five strategic projects with the aim of modernizing financial supervision: open finance, digital supervision, management of new risks, regulatory optimization and institutional modernization.

(More: Eight out of 10 companies in the country are looking to increase their AI budget).

These initiatives are taking significant steps towards modernising financial supervision through the implementation of artificial intelligence (AI), among other models and tools, seeking to improve operational efficiency, increase supervision accuracy and foster innovation in the sector.

As part of this process, The SFC conducted a diagnosis of its technological infrastructure, the capabilities of its staff and the organizational culture to identify strengths and areas for improvement, thereby establishing a solid foundation for AI integration.

In order to process large volumes of data, SFC is implementing a modern data architecture capable of supporting advanced forward-looking models. Some of the key features are: scalability, flexibility to handle large amounts of data and allowing the integration of different types of data sources, both structured and unstructured.

(Read: 63% of employees would use the metaverse, according to a Microsoft survey).

It also has international security standards and maintains privacy and regulatory compliance requirements. This architecture includes Advanced data analysis and visualization tools that enable effective data exploration and analysisThis includes the ability to build predictive and machine learning models to identify patterns and trends in data and make accurate forecasts. Additionally, the new architecture supports rapid iteration and updating of models as data changes and is updated.

Data governance and security are critical elements to ensure that information is managed within the framework of current regulations.

In this line, the SFC is implementing generative AI models for the analysis of financial regulations and the risk of money laundering. With respect to the analysis of financial regulations Machine learning algorithms and natural language processing are being used to analyse large amounts of legal data and extract relevant informationThese algorithms allow the tool to continuously learn and improve as elements of legal analysis are captured based on user feedback.

A collection of machine learning models for digital and online supervision applications (suptech) is also being developed, which will allow the elimination of data reporting currently carried out by financial institutions to meet the information requirements of the SFC. This will allow a significant reduction in the expenses incurred by financial institutions, which sooner or later will result in a reduction in their operating costs and, consequently, in interest rates.

(See: CEOs stopped investing in sustainability due to lack of financial resources).

Within this collection there are real-time data analysis models and a Tool designed to find, extract and compare information that is currently scattered across thousands of articles, technical reports, supervisory reports and banking documents.

With respect to the new risk management project, a forward-looking analysis model is being developed that uses a wide variety of data, such as climate, geospatial, and demographic information, to improve the process of detecting emerging risks. Risk detection is an important component in measuring inherent operational risk.

The entity also has ‘LaArenera’, a controlled environment to develop, test and adjust, among others, AI prototypes in a safe manner. In this way, the SFC can validate the technology, mitigate risks, adjust systems, encourage innovation and train its officials in the use of new technologies. Under the auspices of ‘LaArenera’, a chatbot is being developed to interact with the financial consumer and allow users to easily access current financial regulations.

Regarding the open finance project, The SFC is using a technological ecosystem that promotes transparency and equitable access to financial services by using innovative technologies based on recommendation systems that use generative AIIn this sense, institutional modernization is crucial to implement these changes, improving the efficiency, security and adaptability of the financial system to current demands and promoting a more inclusive and sustainable economy.

(More: Slowing economy to further weigh on big tech rally, Bank of America says)

The Superintendency has a clear roadmap to implement AI with the development of prototypes and the implementation of innovative pilots for the financial sector. It has also established KPIs and metrics to measure the impact of these initiatives, and plans the allocation of resources to ensure that its human talent is prepared to take advantage of AI.

To promote a culture of innovation, the SFC has implemented training programs and created innovation spaces to foster new ideas and technological solutions.

These initiatives improve the efficiency of supervision and position the SFC at the forefront of technological innovation in the financial sector with a strategic focus on facing future trends and ensuring robust and effective financial supervision. In addition, it enables the development of the financial system, making it broader, deeper, less costly and, consequently, with more possibilities for financial inclusion for the benefit of Colombians.

Luis Arregocés (PhD) and Carlos Merlano (DTech candidate)
Office Advisors for Artificial Intelligence
Financial Superintendence of Colombia