OpenAI released a policy paper titled “Industrial Policy for the Intelligence Age: Ideas to Keep People First” in April 2026. The document presents exploratory proposals for governments and companies to manage the transition toward superintelligent AI systems while prioritizing broad economic participation and risk mitigation. It draws parallels to historical industrial shifts and calls for updated tools in taxation, infrastructure, safety nets, and governance to distribute gains and limit downsides.
The paper argues that AI will drive major productivity gains, lower costs for goods and services, and accelerate scientific progress, but without deliberate policy, benefits could concentrate while job displacement and systemic risks intensify. OpenAI positions these ideas as a starting point for discussion rather than a finished blueprint, focusing primarily on the United States with implications for broader coordination.
“ The promise of superintelligence is extraordinary. Just as electricity transformed homes, the combustion engine remade mobility, and mass production lowered the cost of essential goods, superintelligence will speed up scientific and medical breakthroughs, significantly increase productivity, lower costs for families by making essential goods cheaper, and open the way for entirely new forms of work, creativity, and entrepreneurship. ”
Building an Open Economy
The first section focuses on translating AI capabilities into shared prosperity. Key mechanisms include giving workers input on AI deployment to improve job quality and safety, while avoiding uses that erode compensation or intensify workloads. Support for AI-enabled entrepreneurship targets displaced workers through microgrants, standardized contracts, shared infrastructure, and training to lower barriers for small business formation.
A “Right to AI” concept treats access to capable models as essential infrastructure, comparable to electricity or basic literacy. This would require expanded education, connectivity, and affordable foundational models for individuals, small firms, schools, and underserved areas.
Tax policy adjustments address the shift toward capital-intensive production. Options include higher rates on capital gains or corporate income from automated systems, paired with incentives that tie R&D credits to workforce outcomes. Revenue could fund social programs. A public wealth fund, seeded by government and industry contributions, would invest in AI-related assets and distribute returns to citizens.
Energy infrastructure receives direct attention. Public-private partnerships would accelerate grid expansion through investment credits, subsidies with equity participation, and streamlined permitting for high-priority projects, while aiming to limit fiscal risk and share upside.
Efficiency gains from AI should convert into tangible worker benefits such as enhanced retirement contributions, healthcare, or childcare. The paper suggests piloting shorter workweeks with pathways to permanence. Safety nets would become more adaptive, using real-time data on AI-driven disruption to trigger temporary expansions in unemployment support, wage insurance, or retraining, then phasing out as conditions stabilize.
Portable benefits decoupled from single employers would allow continuous accrual of healthcare, retirement, and skills funding across job changes. Investment in human-centered sectors like care work, education, and healthcare could absorb labor shifts, supported by targeted training and wage subsidies. Finally, AI-enabled laboratory networks would speed scientific discovery across universities, hospitals, and regional centers.
Building a Resilient Society
The second section addresses risks from advanced AI, including misuse in cyber or biological domains, potential misalignment, and institutional strain. Proposals center on developing safety systems such as threat modeling, red-teaming, and protective countermeasures with competitive markets encouraged through government procurement and standards.
An “AI trust stack” would incorporate provenance tracking, privacy-preserving audits, and clear accountability lines. Auditing regimes should scale for high-risk models, building on existing institutions while avoiding overreach on lower-risk applications. Containment playbooks would outline coordinated responses if dangerous systems require isolation, modeled on cybersecurity or public health practices.
Corporate governance recommendations favor mission-aligned structures with built-in safeguards against capture. Government use of AI would face codified high-safety standards, while public input mechanisms ensure democratic oversight of alignment goals. Incident reporting, including near-misses, would feed shared learning. International coordination on evaluations and information exchange receives emphasis to manage cross-border risks.
“ We don’t have all, or even most of the answers. Different paths will require different policy responses, and no single set of tools will be enough in any scenario. ”
Investment and Decision Implications
For investors and executives, the proposals signal potential shifts in capital allocation. Energy and grid projects tied to AI compute could see accelerated permitting and blended public-private financing, creating opportunities in utilities, transmission, and supporting supply chains. Tax changes targeting automated production or capital gains may alter after-tax returns on AI-heavy investments, while incentives linked to workforce outcomes could reward firms that invest in reskilling or shared prosperity mechanisms.
Safety and auditing requirements would increase compliance costs for frontier developers but might create markets for specialized testing, verification, and insurance services. Public wealth funds and portable benefits could influence long-term consumption patterns and labor market stability, affecting demand forecasts in consumer and care sectors.
The document does not quantify fiscal costs, revenue potential, or implementation timelines. Historical analogies to electricity and mass production suggest large-scale infrastructure and social contract updates can yield sustained growth when executed effectively, but outcomes depend on execution details, political feasibility, and international alignment.
OpenAI invites feedback via newindustrialpolicy@openai.com and offers fellowships plus API credits to support related work. The paper emphasizes experimentation and iteration over rigid prescriptions.
For capital allocators, the core message is pragmatic: AI capability advances are underway and will reshape production, labor, and risk profiles. Policy frameworks that maintain open access, fund necessary infrastructure, adapt fiscal tools, and build verifiable safety systems reduce downside exposure while preserving innovation incentives. Monitoring legislative responses to these and similar proposals will inform positioning in compute, energy, talent, and defense-adjacent verticals.
The transition to more capable AI systems carries both upside potential in productivity and scientific output and material risks around concentration and disruption. Informed decisions require tracking actual policy development, cost-benefit data as it emerges, and execution track records rather than high-level concepts alone.
The paper argues that AI will drive major productivity gains, lower costs for goods and services, and accelerate scientific progress, but without deliberate policy, benefits could concentrate while job displacement and systemic risks intensify. OpenAI positions these ideas as a starting point for discussion rather than a finished blueprint, focusing primarily on the United States with implications for broader coordination.
“ The promise of superintelligence is extraordinary. Just as electricity transformed homes, the combustion engine remade mobility, and mass production lowered the cost of essential goods, superintelligence will speed up scientific and medical breakthroughs, significantly increase productivity, lower costs for families by making essential goods cheaper, and open the way for entirely new forms of work, creativity, and entrepreneurship. ”
Building an Open Economy
The first section focuses on translating AI capabilities into shared prosperity. Key mechanisms include giving workers input on AI deployment to improve job quality and safety, while avoiding uses that erode compensation or intensify workloads. Support for AI-enabled entrepreneurship targets displaced workers through microgrants, standardized contracts, shared infrastructure, and training to lower barriers for small business formation.
A “Right to AI” concept treats access to capable models as essential infrastructure, comparable to electricity or basic literacy. This would require expanded education, connectivity, and affordable foundational models for individuals, small firms, schools, and underserved areas.
Tax policy adjustments address the shift toward capital-intensive production. Options include higher rates on capital gains or corporate income from automated systems, paired with incentives that tie R&D credits to workforce outcomes. Revenue could fund social programs. A public wealth fund, seeded by government and industry contributions, would invest in AI-related assets and distribute returns to citizens.
Energy infrastructure receives direct attention. Public-private partnerships would accelerate grid expansion through investment credits, subsidies with equity participation, and streamlined permitting for high-priority projects, while aiming to limit fiscal risk and share upside.
Efficiency gains from AI should convert into tangible worker benefits such as enhanced retirement contributions, healthcare, or childcare. The paper suggests piloting shorter workweeks with pathways to permanence. Safety nets would become more adaptive, using real-time data on AI-driven disruption to trigger temporary expansions in unemployment support, wage insurance, or retraining, then phasing out as conditions stabilize.
Portable benefits decoupled from single employers would allow continuous accrual of healthcare, retirement, and skills funding across job changes. Investment in human-centered sectors like care work, education, and healthcare could absorb labor shifts, supported by targeted training and wage subsidies. Finally, AI-enabled laboratory networks would speed scientific discovery across universities, hospitals, and regional centers.
Building a Resilient Society
The second section addresses risks from advanced AI, including misuse in cyber or biological domains, potential misalignment, and institutional strain. Proposals center on developing safety systems such as threat modeling, red-teaming, and protective countermeasures with competitive markets encouraged through government procurement and standards.
An “AI trust stack” would incorporate provenance tracking, privacy-preserving audits, and clear accountability lines. Auditing regimes should scale for high-risk models, building on existing institutions while avoiding overreach on lower-risk applications. Containment playbooks would outline coordinated responses if dangerous systems require isolation, modeled on cybersecurity or public health practices.
Corporate governance recommendations favor mission-aligned structures with built-in safeguards against capture. Government use of AI would face codified high-safety standards, while public input mechanisms ensure democratic oversight of alignment goals. Incident reporting, including near-misses, would feed shared learning. International coordination on evaluations and information exchange receives emphasis to manage cross-border risks.
“ We don’t have all, or even most of the answers. Different paths will require different policy responses, and no single set of tools will be enough in any scenario. ”
Investment and Decision Implications
For investors and executives, the proposals signal potential shifts in capital allocation. Energy and grid projects tied to AI compute could see accelerated permitting and blended public-private financing, creating opportunities in utilities, transmission, and supporting supply chains. Tax changes targeting automated production or capital gains may alter after-tax returns on AI-heavy investments, while incentives linked to workforce outcomes could reward firms that invest in reskilling or shared prosperity mechanisms.
Safety and auditing requirements would increase compliance costs for frontier developers but might create markets for specialized testing, verification, and insurance services. Public wealth funds and portable benefits could influence long-term consumption patterns and labor market stability, affecting demand forecasts in consumer and care sectors.
The document does not quantify fiscal costs, revenue potential, or implementation timelines. Historical analogies to electricity and mass production suggest large-scale infrastructure and social contract updates can yield sustained growth when executed effectively, but outcomes depend on execution details, political feasibility, and international alignment.
OpenAI invites feedback via newindustrialpolicy@openai.com and offers fellowships plus API credits to support related work. The paper emphasizes experimentation and iteration over rigid prescriptions.
For capital allocators, the core message is pragmatic: AI capability advances are underway and will reshape production, labor, and risk profiles. Policy frameworks that maintain open access, fund necessary infrastructure, adapt fiscal tools, and build verifiable safety systems reduce downside exposure while preserving innovation incentives. Monitoring legislative responses to these and similar proposals will inform positioning in compute, energy, talent, and defense-adjacent verticals.
The transition to more capable AI systems carries both upside potential in productivity and scientific output and material risks around concentration and disruption. Informed decisions require tracking actual policy development, cost-benefit data as it emerges, and execution track records rather than high-level concepts alone.