Bill Gates warns AI job losses could undermine taxes and social safety nets

The erosion of the tax base

Bill Gates warns that artificial intelligence could fundamentally alter the relationship between employment and public finances. As companies adopt automation, the current incentive structure may favor replacing human workers with machines to maximize efficiency. This shift threatens to weaken the tax revenues that governments rely on to provide social protections.

Most developed economies fund public services through taxes on human labor and individual income. If AI replaces a significant portion of the workforce, these revenue streams could shrink even as the overall economy becomes more productive. Gates notes that the revenue for the safety net will actually be reduced as prosperity improves, creating a difficult fiscal gap for policymakers.

The challenge lies in the fact that productivity gains from AI might not translate into higher tax collections. Governments may find themselves with fewer resources to help displaced workers find new roles or gain necessary skills. This situation represents a significant test for the quality of government in managing a technological transition.

Proposed solutions for the shift

To address the potential loss of tax revenue, Gates has proposed a specific "token tax" on artificial intelligence. This model would link taxes to the amount of data and information handled by AI systems. By taxing the usage of the technology itself, the government could capture a portion of the value generated by automation.

In addition to taxing AI usage, Gates advocates for raising taxes on the world's wealthiest individuals. He suggested that governments need to find a way to tax very wealthy people in a manner that feels fair to the public. He noted that while higher taxes on billionaires are necessary, they would likely not be enough to solve all existing fiscal pressures.

Gates shared that he has contributed roughly $14 bn in taxes over his life but would have been willing to contribute $40 bn. He acknowledges that because government spending often exceeds tax revenues in many countries, more taxes will eventually have to be raised. At the same time, some government spending will likely have to be curbed to balance the budget.

Widening wealth inequality

Research from the International Monetary Fund (IMF) published in 2024 supports the concern that AI benefits may be distributed unevenly. While the technology could lead to substantial productivity improvements and higher incomes for many, the distribution of those gains remains uncertain. The outcome depends heavily on how AI changes the demand for different types of labor across the economy.

If AI primarily complements the work of higher-paid employees, it could lead to an increase in labor income inequality. Furthermore, the IMF suggests that greater returns to capital could widen the gap between asset owners and workers. This aligns with Gates' warning that the financial benefits of automation may accrue disproportionately to companies and owners of capital.

For advanced economies, the IMF identifies policies that support employment transitions as a top priority. Protecting workers who are adversely affected by AI is essential to maintaining social stability. Without proactive intervention, the transition could leave displaced employees struggling to adapt to a rapidly changing labor market.

The scope of AI automation

Gates describes the rise of advanced AI as an "evolutionary event" where people are no longer the most intelligent beings on Earth. He expects AI to eventually replace almost all categories of human work skills. This includes the skills required for invention and white-collar work, as well as physical human labor through the use of humanoids.

The speed and scope of this transition are what make it a unique challenge for modern society. While AI offers significant scientific benefits, such as improving the process of drug discovery, the social consequences are profound. Managing these changes requires a level of coordination that many governments may find difficult to achieve in the current climate.

The timing of this shift is particularly difficult because many rich countries are already facing heightened geopolitical tensions and budgetary pressures. Gates suggests that this is not the best time for nations to attempt the hardest thing they have ever done. However, the move toward automation appears inevitable, making the discussion around taxation and social support more urgent.