Earth at night (NASA)

Post-AGI Economic Safety Forum

White Paper | November 2026

The Post-AGI Economic Safety Forum is being formed around a single question: how do we keep an economy built on human labor safe for the people who live in it when labor is no longer scarce? This paper states the problem, defines what we mean by economic safety, and sketches the policy directions the forum is intended to develop.

1. The Premise

Artificial intelligence is becoming a general-purpose substitute for human cognitive work. Reasonable people disagree about how far and how fast this substitution will go. What is no longer seriously contested is the direction: each year, a wider range of tasks that used to require a person can be done by a machine at lower cost.

Most economies distribute income primarily through wages. Most public budgets are funded primarily by taxing labor and consumption. Most households hold little capital and depend on a paycheck. This arrangement works when human labor is the scarce input to production. It is not designed for a world in which it is not.

The safety problem is therefore not only whether advanced AI systems behave as intended. It is whether the economic institutions that stand between people and destitution can adapt as quickly as the technology displaces them.

2. What Economic Safety Means

We use economic safety to mean three conditions that should hold before, during, and after any transition to highly automated production:

• Broad-based income security. Every household retains a reliable claim on a share of what the economy produces, whether or not it can sell labor.
• Broadly held ownership. The returns to the machines doing the work are not concentrated so narrowly that economic and political power become the same thing.
• Resilient public finance. Governments can fund services and safety nets even if payroll and income taxes shrink as a share of the economy.

Economic safety is not an argument against automation. Abundance created by machines is the best plausible outcome, if its gains are widely shared. The concern behind this forum is the gap between that outcome and the institutions we actually have.

3. Why Existing Institutions Lag

Twentieth-century economic safety nets assume employment is the normal state of an adult and that wages are the main thing to tax. Unemployment insurance pays out when you lose a job, not when the job category disappears. Pensions and retirement accounts compound over a working life. Healthcare, in many countries, is tied to an employer. Corporate and antitrust law evolved for markets in which scale was checked by the cost of people, plants, and logistics.

Each of these institutions can, in principle, be reformed. But institutional reform is measured in years and decades, and capability gains are now measured in months. The mismatch in speeds is the core risk: not that automation arrives, but that it arrives before the institutions that would make it safe.

4. Directions Worth Developing

The forum is not meant to endorse a single program. Its purpose is to develop and pressure-test a portfolio of mechanisms that could, in combination, deliver economic safety:

• Broadening ownership of productive capital. Sovereign and public investment funds, citizen shareholding, and dividend mechanisms that give every person a direct stake in automated production.
• Taxing economic rents rather than labor. Levies on compute, on excess returns to AI capital, and on land and other fixed factors, so that the public share of production does not depend on taxing wages.
• Income floors and transition insurance. Wage insurance, portable benefits, and basic income designs that decouple survival from continuous employment without destroying the incentive to contribute.
• Competition and access policy. Rules that keep frontier AI capability, compute, and data from consolidating into a small number of private chokepoints.
• Measurement. Better, faster data on which tasks and occupations are actually being automated, so policy responds to evidence rather than vibes.

5. What the Forum Is For

The forum is intended as a working group, not a think tank with a finished doctrine. The plan is to convene economists, engineers, policymakers, and practitioners; publish short working papers like this one; and maintain open questions where the honest answer is that nobody knows yet.

The proposed near-term agenda is unglamorous: map the tax and benefit systems most exposed to labor displacement, quantify the fiscal exposure, identify which mechanisms are deployable at which scales, and say clearly what would change our minds. Economic safety is an engineering problem wearing institutional clothes, and it deserves to be worked on with that seriousness.

Post-AGI Economic Safety Forum. Working draft, November 2026.