Sections

Research

AI’s economic peril to democracy

Stephanie A. Bell and
Stephanie A. Bell Chief Programs and Insights Officer - Partnership on AI
Anton Korinek

March 14, 2024


Key takeaways

  • AI-driven inequality threatens democracy through elite influence and discontent.
  • Policies should guide AI to augment workers and distribute the gains.
  • Inclusive reforms can strengthen democracy against AI-fueled inequality.
  • If AI-driven job disruption rises, alternative income distribution may be needed.
Robotic hand touching wooden blocks with images of ethics and laws
Shutterstock / 3rdtimeluckystudio
Editor's note:

This article is an executive summary of a report published by the Journal of Democracy. You can read the pre-publication manuscript here.

Executive Summary

Artificial intelligence (AI) is poised to fundamentally reshape labor markets and the economy. If the gains accrue narrowly while human labor loses value, economic inequality could substantially worsen. In our recent article on “AI’s Economic Peril” for the Journal of Democracy, we argue that this poses a grave threat to democracy that is separate from more traditional AI risks to democracy such as deep fakes and misinformation: High inequality corrodes democratic institutions through increased elite influence, corruption, populism, and greater public discontent. At the same time, weakened democracy loses power to reign in inequality through progressive policies.

This may create a vicious feedback loop of eroding democracy and rising inequality, which may accelerate rapidly following an economic shock like large-scale displacement of workers by AI. The result could be a new society-wide equilibrium with starkly increased income disparities and a weakened voice for ordinary citizens.

To avert this vicious cycle, policy interventions should pursue two complementary goals: 1) Mitigate inequality arising from AI’s economic impacts, and 2) Strengthen democracy against the corrosive effects of increased inequality.

Several policies show promise for reducing the risk that AI may increase inequality:

  • Guide AI development to create new opportunities for humans and augment their abilities, rather than solely automating work. Governments can encourage this through R&D funding rules, while companies can make human-centric AI part of their technology strategy.
  • Empower workers through promoting unionization, collective bargaining, co-determination of technology, and other means of giving labor voice and leverage. This is especially important in white-collar fields likely to be disrupted.
  • Reform tax policies so that hiring humans is not disadvantaged compared to acquiring AI systems and automation. This removes incentives to replace workers.
  • Use antitrust enforcement and regulation to prevent excessive concentration of power and gains from AI development. Authorities should be vigilant about monopolization.

To directly strengthen democracy, inclusive political reforms are needed, including:

  • Improving electoral integrity through transparency, security, access, and representation. This bolsters citizens’ political voice.
  • Reducing the outsized influence of wealthy interests on the policy process through reforms. This could include public campaign financing and closing “revolving doors” between government and industry.
  • Building more participatory institutions that empower ordinary citizens, such as participatory budgeting whereby citizens have the opportunity to allocate parts of the public budget through a deliberative process. This gives people more direct say over policies.
  • Designing AI systems themselves to depolarize public discourse rather than contribute to tribalism. This can facilitate broad-based policymaking.

If AI advances eventually create technological unemployment, more sweeping reforms may become necessary. Policymakers should avoid short-sighted work requirements as a precondition for receiving social services. Instead, if labor markets are undermined, alternative mechanisms for distributing income equitably may be needed. With thoughtful policies, however, a future of shared prosperity alongside AI advances remains achievable. Although challenging, the path to prevent inequality from weakening democracy still lies open before us.

Read the full report

Authors

  • Acknowledgements and disclosures

    The Brookings Institution is financed through the support of a diverse array of foundations, corporations, governments, individuals, as well as an endowment. A list of donors can be found in our annual reports published online here. The findings, interpretations, and conclusions in this report are solely those of its author(s) and are not influenced by any donation.