Annual Economic Report 2024, Chapter III: Artificial intelligence and the economy: implications for central banks

Claims Supported by This Source

1. Finance is one of the sectors with the greatest opportunities and risks from AI. The impact is demonstrated in four areas: payments, lending, insurance, and asset management.
2. AI increases efficiency in back-office operations, regulatory compliance, fraud detection, and customer service. Prime examples are improvements in KYC (Know Your Customer) and fraud detection.
3. In lending, creditworthiness can be assessed using alternative data, such as bank account transactions and payment records for rent, utilities, and communication fees, as well as non-financial data like educational background.
4. In a survey cited by the BIS, about 70% of financial services firms worldwide use AI for cash flow forecasting, credit scoring, and fraud detection.
5. New risks: proliferation of cyberattacks such as phishing; bias and discrimination (entrenching disparities in access to credit); dependence on a few AI providers; and financial stability risks from reliance on the same algorithms amplifying herding behavior, liquidity hoarding, bank runs, and fire sales.
6. In a representative survey of U.S. households, trust in generative AI is lower than in services handled by humans in critical areas like banking and public policy.
7. Central banks can also use AI for compiling statistics, analysis, supervision, and payment system oversight. In the BIS's Project Aurora, machine learning outperformed rule-based money laundering detection.

How to read the verification levels. "Read Through" means the source was read in its entirety. "Bibliography & Abstract Read" means the bibliographic information and abstract were confirmed. "Via Course Material" means the source was referenced through the textbook or course text. The "Text Hash" is the SHA-256 of the main text from the publisher's page, excluding tags and whitespace, used to detect if the source has been modified.