The political economy of AI-driven financial supervision

As the financial services industry and the authorities overseeing it begin to turn to Artificial Intelligence (AI), what social benefits and costs will AI bring to the industry? What issues will arise as some financial centres and firms embrance AI, as others wait for regulation?

In his working paper, 'The Political Economy of AI-Driven Financial Supervision' Prof. Gerard Hertig of Future Resilient System (FRS) programme, surveys and identifies the impact of AI applications on the supervision of financial services.

In short, the use of AI should improve risk management and labour productivity across the financial services industry. AI is also favoured for its more complete and unbiased decision-making. In ‘normal times’, AI is expected to both reinforce private sector autonomy and decrease supervisory enforcement costs. When it comes to ‘extreme events’ (i.e. when dealing with systemic risk), AI-use will neither facilitate their supervisory handling nor increase systemic risk.

Prof. Hertig predicts that that AI self-regulation is likely to become dominant in normal times, whereas state regulation will (slowly but increasingly) target systemic issues.

Adoption of AI, however, also comes with social costs. AI-use by a small handful of technologically-advanced countries and firms will contribute to a digital gap, and may probably allow these intermediaries to game the regulatory system. AI-use is also expected to result in significant job losses, and the public pressure that comes with such disruptions.

For full details, please download the working paper "The Political Economy of AI-​Driven Financial Supervision" (PDF, 233 KB).