SAS offers AI-assisted software for banking, insurance

Naeem Sidiqi, senior SAS advisor of SAS Institute
Naeem Sidiqi, senior SAS advisor of SAS Institute

The Thai financial and insurance sectors have turned to artificial intelligence (AI) and advanced  analytics to manage increased credit risks amid the country’s growing  household debts so as to avoid more non-performing loans, according to a senior executive of SAS Software.


Nutapone Apiluktoyanunt, managing director, SAS Software (Thailand), said US-based SAS, which has been in the Thai market for over 23 years,  has helped Thai banks and insurance companies increase efficiency in credit risk management using AI and advanced analytics..


SAS is also a leader in identity and digital fraud analytics, allowing organizations to detect and adapt to fraud trends via anomaly detection, AI and machine learning.   


Naeem Sidiqi, senior SAS advisor of SAS Institute, said climate risks have become more prominent, especially with regard to more incidents of cyclones, floods, and draughts, among others, around the world.


As a result, banks will soon provide loans based on local weather conditions in each geographical area to manage their climate risks, while green asset ratios will be used to comply with local and international regulations on environmental and social governance.


For example, Thai banks such as Kasikorn Bank may require clients such as Thai Airways to devise a de-carbonisation plan as part of a loan package while lendings to oil companies will be regarded as “brown” assets because they contribute to climate change.


On the Thai government’s plan to grant three new virtual bank licenses,  Nutapone said new operators will likely target potential customers who are currently under-banked and unbanked so that they have better credit access.


Virtual banks will have no physical branches and use new IT infrastructure to tap customers and  manage their credit risks,  providing faster loans at a lower operating cost. Data will be key to decision making on providing loans so data quality is important.


According to Naeem, data from mobile phone bills or monthly electricity charges are reliable but he cautioned that data from most social media are unreliable for credit risk assessment. 


On the role of generative AI,  he said, there is still a lot of hype as far as credit risk assessment is concerned because the current technology is still not good at contextual clues.


However, the issuing of virtual bank licenses will help expand the country’s GDP by providing more credit access, especially micro loans, to a large population at a low cost, while also offering options to current bank customers who are not satisfied by today’s products. 


For example, the co-called virtual banks may offer higher interest rates on saving accounts  than those offered by traditional banks because they have lower operating costs.