In U.S. politics, discussion is rising over an “AI dividend” as a way to respond to job changes stemming from the spread of artificial intelligence (AI).

On April 21 local time, Cointelegraph reported that New York state assemblyman Alex Bores unveiled a policy concept to make direct cash payments to U.S. citizens in preparation for job displacement caused by AI.

Bores said on X, formerly Twitter, that the United States must prepare for the possibility of large-scale labor replacement by AI. He described the plan as an “AI economic response plan.” He outlined securing funding through tax-system changes, returning it to citizens in the form of a dividend, and encouraging companies to keep human labor rather than use AI.

At the core of the policy is creating an institutional mechanism to ensure the gains from higher productivity do not concentrate only in a small number of companies. Under the proposed structure, if AI actually replaces labor, some of the related economic benefits would be distributed directly to citizens. Funding options include taxing AI use, securing stakes in major AI companies, and tax changes for labor and capital.

The dividend funding would not be used only for cash payments. Bores said it would also be spent on job transitions and re-education, skills training, and building AI oversight and safety infrastructure. The policy document said, “If AI dramatically boosts productivity and concentrates wealth, citizens should also have a stake in those gains.”

The proposal comes as concerns grow about how AI’s spread could affect the labor market. A recent Goldman Sachs report estimated that about 16,000 jobs have been lost each month over the past year due to AI adoption. Major tech companies including Amazon, Meta, Intel and Microsoft are reported to have carried out, or to be considering, workforce restructuring citing AI-driven efficiency.

Some urge caution over the view that AI will immediately lead to mass unemployment. Morgan Stanley, in a recent report, assessed that AI’s impact on the labor market has been limited so far, and cited cases in which past technological innovation expanded employment over the long term. It added that the possibility AI could follow a different path cannot be ruled out.

In this environment, Bores’ AI dividend concept is seen as a policy experiment focused on who should receive productivity gains from the spread of AI and how, and how to design a social safety net for automation shocks. Separate from whether it is institutionalized, attention is on whether discussion in U.S. politics will intensify on taxing AI use and linking the proceeds to citizen returns and support for labor transitions.