In a recent speech at an AI summit in Switzerland, IMF First Deputy Managing Director Gita Gopinath cautioned that while artificial intelligence (AI) offers numerous benefits, it also poses grave risks that could exacerbate economic downturns. Gopinath emphasised that while discussions around AI have predominantly centred on issues like privacy, security, and misinformation, insufficient attention has been given to how AI might intensify economic recessions.
Historically, companies have continued to invest in automation even during economic downturns. However, Gopinath pointed out that AI could amplify this trend, leading to greater job losses. According to IMF research, in advanced economies, approximately 30% of jobs are at high risk of being replaced by AI, compared to 20% in emerging markets and 18% in low-income countries. This broad scale of potential job losses could result in severe long-term unemployment, particularly if companies opt to automate jobs during economic slowdowns to cut costs.
The financial sector, already a significant adopter of AI and automation, faces unique risks. Gopinath highlighted that the industry is increasingly using complex AI models capable of learning independently. By 2028, robo-advisors are expected to manage over $2 trillion in assets, up from less than $1.5 trillion in 2023. While AI can enhance market efficiency, these sophisticated models might perform poorly in novel economic situations, leading to erratic market behaviour. In a downturn, AI-driven trading could trigger rapid asset sell-offs, causing market instability. The self-reinforcing nature of AI models could exacerbate price declines, resulting in severe asset price collapses.
AI's integration into supply chain management could also present risks. Businesses increasingly rely on AI to determine inventory levels and production rates, which can enhance efficiency during stable economic periods. However, Gopinath warned that AI models trained on outdated data might make substantial errors, leading to widespread supply chain disruptions during economic downturns. This could further destabilise the economy, as inaccurate AI predictions might cause supply chain breakdowns.
To mitigate these risks, Gopinath suggested several strategies. One approach is to ensure that tax policies do not disproportionately favour automation over human workers. She also advocated for enhancing education and training programs to help workers adapt to new technologies, along with strengthening social safety nets, such as improving unemployment benefits. Additionally, AI can play a role in mitigating its own risks by assisting in upskilling initiatives, better targeting assistance, and providing early warnings in financial markets.
Gopinath accentuated the urgency of addressing these issues, noting that governments, institutions, and policymakers need to act swiftly to regulate AI and prepare for labour market disruptions. Her call to action comes as a reminder that while AI holds great promise, its potential to deepen economic crises must be carefully managed to protect global economic stability.
