Who should benefit from AI productivity?
AI can help people perform some tasks faster. How the gains are shared depends on bargaining power, competition, ownership and public policy.
By the Rethinkist · 8 November 2026 · Published by OM Group Limited
Productivity gains are possible. Distribution is a separate question.
Productivity describes how much output is produced with a given amount of input. An AI assistant might help with writing, coding or customer support, but faster task completion is not the same as higher wages, fewer working hours or better outcomes for everyone.
In a study of customer-support agents, Brynjolfsson, Li and Raymond found that access to a generative AI assistant increased issues resolved per hour on average, with gains varying substantially by experience and skill. This is evidence from a particular workplace and tool, not a forecast for every occupation. See the NBER research paper and published-version reference.
Research about time saved cannot, by itself, tell us how much income will shift between workers, businesses and customers. Adoption costs, errors, oversight, energy use and work reorganisation all matter. The answer also contains a value judgement: what should happen, rather than only what might happen.
Five ways to think about sharing the gains
Workers: pay, time and security
Workers contribute judgement, knowledge and the work required to integrate new tools. A larger share could mean higher pay, shorter working hours without a pay cut, training or stronger job security. Whether workers receive those benefits depends partly on negotiation and employment institutions.
Customers: prices and access
Lower production costs can create room for lower prices or improved services. Competition may encourage businesses to pass some savings on. That is a possible route, not an automatic outcome, especially where customers have few alternatives.
Businesses and investors: returns and reinvestment
Companies fund tools, infrastructure and organisational change and bear risks when deployments fail. Retained returns can support investment. How much return is reasonable, and how much results from market power, remains open to scrutiny.
Society: public services and shared capacity
Public research, education and infrastructure help make innovation possible. Tax systems and public investment can channel some gains toward healthcare, education, social protection or climate-related needs. Choices about design, incentives and international coordination involve trade-offs.
It depends: context and combinations
The best allocation may differ between a public hospital, a small firm and a dominant technology platform. Several groups can benefit at once. Our poll asks who should benefit most, so it deliberately simplifies a more complex distribution question.
What could change the outcome?
- Worker voice and bargaining over pay, workload and time.
- Competitive markets and accountability for quality and prices.
- Tax and spending choices about public benefits.
- Access to useful tools, skills and infrastructure.
- Measurement of quality, safety, energy use and other costs alongside speed.
These are mechanisms to examine, not a claim that one policy will work everywhere. Public choices and business decisions shape who can participate and who absorbs the costs.
Read the wider story
Moana Earth: AI Could Change Everything. But for Whom? provides related editorial context on AI and its wider consequences. Disclosure: Moana Earth and Rethinkist share the operator OM Group Limited. This link is not independent corroboration of our polling results.
What do you think should happen?
The evidence informs the discussion. Your answer expresses your priorities. This voluntary poll cannot establish a representative public view.
Add your voice ↗