AI boom poses new financial stability risks, BIS head says

AI's rapid growth is introducing new financial stability risks, according to the head of the Bank for International Settlements, Pablo Hernandez de Cos. The bank's chief explains that infrastructure spending related to AI has reached a scale that can impact global economic conditions. Central banks face challenges interpreting economies affected by AI, as it influences demand, supply, and financial markets concurrently.

The BIS estimates that the world's five largest technology firms will invest over $1 trillion in AI between 2025 and 2026, with global AI investments projected to rise from around $500 billion to a potential $4 trillion by 2030.

Hernandez de Cos emphasizes that AI's long-term impact hinges on policy decisions, investment in skills and infrastructure, and equitable benefit distribution. He notes that the boom in AI is increasingly fueled by debt and private credit rather than corporate earnings, which warrants careful examination due to its opaque and interconnected nature.

Additionally, AI is reshaping global trade flows, benefiting economies such as South Korea, Singapore, Malaysia, and Taiwan by strengthening their export prices for AI chips and equipment.

Studies indicate that generative AI can significantly enhance productivity, with gains ranging from 10% to 65% in specific tasks, particularly coding, consulting, and professional writing. However, the extent to which these improvements translate into overall economy-wide productivity growth remains uncertain. Estimated gains in total factor productivity growth due to AI are projected at around half a percentage point per year, contingent on the rate of adoption and effective reallocation of labor and capital.

Advanced economies are anticipated to gain first due to their larger service sectors and higher readiness to adopt AI. Emerging economies, on the other hand, face varied prospects, but Hernandez de Cos believes India has a genuine opportunity to reduce the gap, aided by its digital public infrastructure. While AI can increase workers' productivity, it also has the potential to replace routine cognitive tasks, leading to job losses in sectors like customer service, programming, and administration. As such, retraining and reskilling become crucial to mitigate the impact of these changes.

Lastly, Hernandez de Cos warns that lofty valuations, market concentration, and opaque financing structures could create vulnerabilities if corporate profits fail to meet expectations. He does not claim that this is where the AI boom must lead, but the scale and speed of current investment and the anticipated commercial returns warrant caution, drawing parallels with previous booms such as the railway expansion era and the dotcom surge.