Tunisia's Ministry of Finance and the General Directorate of Customs are planning to implement artificial intelligence (AI) and machine learning technologies to improve risk management and monitoring. The new system will enable the analysis of large amounts of data to identify high-risk files and permits, allowing for more targeted monitoring and control. This approach aims to prioritize efforts and optimize resource allocation.
The project was presented during a national meeting organized by the National School of Finance and the National School of Customs. The meeting focused on the challenges and opportunities of AI in the fields of finance, taxation, and customs. According to Essam Al-Fasatwi, Director of Risk Management at the General Directorate of Customs, the AI model has shown promising results, detecting additional amounts of around 96.9 million dinars with a detection rate of 49.23%.
The AI model used in the experiment relied on 11 million observations and 27 initial variables, which were enriched with composite variables to reach 42 variables. The model was then reduced to 11 variables to build the final model. The "decision tree" model was chosen for its interpretability, allowing for clear explanations of the reasons behind classifying a permit as high-risk.
The customs administration's experience relies on integrating monitoring results into the model's retraining and periodic updates. This approach enables a shift from static selectivity based on predefined rules to dynamic selectivity that benefits from data and previous monitoring results. In the tax field, the General Director of the Information Center at the Ministry of Finance, Saleh Al-Madab, presented a project to use big data and machine learning to improve tax monitoring.
The project aims to identify high-risk files and direct resources towards them, while also enhancing the administration's ability to monitor the informal economy and new forms of digital activity. The project is currently in the bidding phase, with an expected implementation period of one year. The technical infrastructure will be based on local architecture and open-source tools.
According to Moez Deldoul, General Director of Performance at the General Directorate of Performances, digitalization has reduced physical and repetitive tasks, freeing up time for monitoring and analysis. He highlighted that the number of remote declarants has reached 157,000, while the number of landlord permits deposited within deadlines has increased to 87,000, up from 57,000 previously.
The use of AI is expected to improve service quality, verify data consistency, and link transactions between different stakeholders. It will also enable the ranking of files according to risk level, allowing controllers to focus on audit, investigation, and analysis tasks. However, this transformation requires the development of new skills among agents, particularly in statistics, data analysis, and understanding model outputs.
Key points
- The implementation of AI and machine learning in Tunisia's customs and finance ministry aims to improve risk management and monitoring.
- The new system will enable the analysis of large amounts of data to identify high-risk files and permits.
- The project is expected to enhance the administration's ability to monitor the informal economy and new forms of digital activity.
