How will an anti-fraud system founded in artificial intelligence and machine learning make a difference in Morocco?
Rabat – The Spanish-based credit risk management company, AIS Group, has started implementing Morocco’s first anti-fraud system founded in machine learning and artificial intelligence through the Moroccan company Salafin.
Salafin finances consumer credit and the purchases of vehicles for individuals, as well as small and medium-sized enterprises. It is also a subsidiary of BMCE, Banque Marocaine du Commerce Extérieur. Salafin detected 700 confirmed fraud cases in 2017.
Daniel Torrents, the head of business and development for AIS Group in Morocco and Africa, described the system’s fraud-prevention strategy. “A model has been put in place that calculates the likelihood of fraud in credit applications filed with financial institutions. Thus, the risk of fraud is evaluated before the final loan is granted,” stated Torrents.
Torrents elaborated that the artificial intelligence techniques “have a much greater predictive ability than traditional techniques and, as a result, a higher success rate detecting fake accounts, including non-existent businesses.” Additionally, “credit assessment and fraud detection models combine credit risk and operational risk to integrate a single view of the transaction and facilitate overall risk control.”
AIS Group predicts that in Moroccan financial institutions, the artificial intelligence based system should improve fraud detection by 20 to 30 percent. Salafin’s project is the first model to use the machine learning method in Morocco.
In 2018, the international security company McAfee released a reported that detailed the economic impact of cybercrime worldwide. Globally, in 2017, the cost of cybercrime was around $608 billion and in the MENA region, around $5 billion.
The implementation of artificial intelligence to curb cybercrime will significantly reduce economic loss, according to Forbes magazine. Artificial intelligence is capable of learning relationships in modeled data, in turn identifying fraudulent activities and preventing them from causing harm.