Former CB Governor urges AI credit models to democratise SME funding

• Says emergence of AI-driven credit mechanisms offers a profound paradigm shift in how risk is assessed, effectively bypassing outdated physical asset requirements
• Urges financial sector to drastically elevate its focus on data protection and cybersecurity

​By Nishel Fernando

​Sri Lanka’s private sector credit is expanding at a robust pace, but effectively directing these funds into the real economy, specifically to the small and medium enterprises that form its structural backbone, remains a stubborn policy challenge. 

While recent data points to a strong rebound in private sector credit, a deeper economic analysis reveals a systemic concern.

If this renewed credit cycle disproportionately fuels consumption-led imports rather than domestic production, value-added exports, and technology, the country risks diluting its hard-won macroeconomic stabilization. 

Former Central Bank Governor Dr. Indrajit Coomaraswamy suggests the solution lies in a technological paradigm shift, advocating for the urgent adoption of artificial intelligence in risk assessment to bridge this critical credit gap.

​Speaking at the Sri Lanka Institute of Directors CEO Forum, Dr. Coomaraswamy highlighted the structural flaws in the current lending framework. 

“As we look at modernising our financial architecture, we cannot ignore the role of technology in democratizing access to capital,” he stated. 

For decades, the rigid reliance on traditional collateral has systematically marginalized the SME sector and agricultural entrepreneurs, keeping them locked out of the formal banking system and stifling inclusive growth.

To counter this, he argued that the emergence of AI-driven credit mechanisms offer a profound paradigm shift in how we assess risk,” effectively bypassing outdated physical asset requirements.

​By leveraging alternative digital footprints—ranging from mobile payment flows and supply chain data to utility payment histories—these AI models can build highly accurate, dynamic risk profiles for borrowers who entirely lack formal credit histories. This technological leap allows banks to accurately price risk and disburse loans based on actual cash flow metrics and operational viability.

If the Central Bank and local financial institutions can properly integrate these AI credit scoring models within a robust regulatory sandbox, it would not only accelerate financial inclusion but also direct credit far more efficiently into the productive, growth-driving sectors of our economy, Dr. Coomaraswamy explained.

​However, modernizing the financial architecture with alternative data sources naturally introduces a new spectrum of systemic vulnerabilities. Acknowledging this shifting risk landscape, Dr. Coomaraswamy emphasised an urgent need for the financial sector to drastically elevate its focus on data protection and cybersecurity. As AI models consume vast amounts of alternative digital information to score SMEs, defending against cyber threats must become a continuous, board-level priority. 

He stressed that without establishing robust cyber resilience and stringent data privacy controls alongside these credit models, the fundamental trust required to sustain a digitised financial ecosystem would fracture, undermining the very technological advancements meant to drive Sri Lanka’s economic recovery.