Taxing AI's Winners to Insure its Losers

Key Takeaway

AI and automation will likely lead to increases in income inequality. Governments need to implement "inequality insurance" by automatically increasing tax rates on the top 1% when inequality exceeds certain thresholds. This will help compensate workers who lose jobs to automation.


  • AI and automation will dramatically increase productivity but also pose risks of mass unemployment and sharply increased income inequality as owners of technology amass wealth.
  • Regulation alone cannot prevent these negative impacts without also limiting potential benefits of AI. So policies are needed to compensate citizens if disaster occurs.
  • As AI economist example shows, tax rates on top 1% could automatically increase if their share of national income rises above current levels. Goal is stopping further increases in inequality.
  • Such "inequality insurance" built into tax system could fund minimum basic income or wage subsidies for displaced workers.
  • Full insurance might not be possible in most extreme scenario where top 1% receive all pretax income. But partial insurance still warranted and politically more feasible than capping incomes.
  • Setting higher inequality caps and expanding beneficiary class beyond just top 1% would also increase political viability.
  • Explicitly defining adaptive tax rates now is better idea than in past given AI's potential to massively accelerate inequality. Governments must prepare insurance policies now.