Speaking at the 90th High-Level Conference organized by the RBI in New Delhi, Governor expressed concern that the growing use of AI could enable a few technology providers to dominate the market, creating systemic vulnerabilities.
- Currently, AI in Financial System is being used in Algorithmic and high-frequency trading, credit scoring and approvals, customer services through instruments like chatbots, predictive analytics for market trends for risk management, etc.
Risks posed by AI to banking and financial services
- Concentration risks: If many financial institutions use similar AI models for trading or risk assessment, a failure or error in these algorithms can have cascading effects across global financial markets.
- e.g., AI trading systems can amplify market volatility by triggering mass sell-offs during downturns.
- Algorithmic biases: AI systems are trained on historical data can lead to unfair practices like discriminatory lending or credit decisions.
- e.g., An AI-driven loan approval system may inadvertently deny loans to specific demographic groups.
- Data security and privacy: Breach or misuse of data can lead to identity theft, fraud, and significant losses for both institutions and customers.
- Others: Lack of transparency due to ‘Black Box’ problem, misleading information due to ‘AI hallucinations’, etc.
Measures to be taken to address these risks
|
