In a new report, the International Monetary Fund (IMF) underscores the pivotal role of fiscal policies in ensuring equitable distribution of benefits from generative artificial intelligence (AI). The report highlights AI’s potential to enhance productivity and transform public services like healthcare and education. However, it also warns of AI’s rapid transformation posing risks of labour market disruption and increased inequality.
To navigate these challenges, governments are urged to adopt agile fiscal policies that anticipate and mitigate potential disruptions. Stronger unemployment insurance and re-skilling programs are recommended to cushion short-term job losses caused by AI. Social safety nets need adaptation to manage prolonged unemployment scenarios resulting from AI adoption.
The IMF report reveals that 40% of global employment is susceptible to AI impacts. Moreover, advanced economies are facing even higher exposure. Conversely, emerging markets and low-income countries are less prepared to manage AI-induced transitions. A big reason for this flexibility is the larger populations outside of traditional employment or educational systems.
Addressing concerns about taxing AI, the report advises against specific AI investment taxes that could hinder productivity. Instead, it proposes strengthening existing tax systems to ensure fair taxation of capital income, potentially narrowing global wealth disparities. Recommendations include a global minimum wealth tax, supplementary taxes on excess profits, and enhanced taxes on capital gains.
The IMF authors emphasize that current fiscal decisions will shape AI’s impact for decades to come. By implementing prudent fiscal policies, governments can harness AI’s potential for broader societal benefits while minimizing adverse effects on income inequality and labour markets.
Overall, the report advocates for proactive measures to harness AI’s transformative potential responsibly, ensuring it delivers sustainable gains across global economies and societies.
About Author
