OpenAI on Monday published a set of policy proposals designed to tackle what it says will be the major social and economic disruptions from ever-more powerful artificial intelligence, urging governments to consider sweeping changes to taxes, benefits and regulation before the changes become entrenched.
The package, which OpenAI characterizes as “ambitious” and “intentionally early and exploratory,” lays out recommendations ranging from a new industrial policy agenda to a rethinking of tax bases, portable healthcare and retirement accounts, and “common-sense” rules for AI governance. The company frames the proposals as preventative: unless public policy keeps pace with technological change, it warned, existing institutions and safety nets risk being left behind as AI shifts production away from labor and toward capital.
Central to OpenAI’s argument is the expectation that AI-driven efficiency gains will shrink the role of labor in some sectors, shifting more economic returns to corporate profits and capital gains. That shift, the company said, could erode payroll-tax revenue streams that fund Social Security, Medicaid, SNAP benefits and housing assistance. To counter that risk, OpenAI suggests policymakers consider higher taxes on capital gains or corporate income, or targeted levies on “sustained AI-driven returns” — proposals the company said would help maintain funding for social programs even as employment patterns change.
OpenAI also urged policymakers to make benefits portable rather than employer-tied. It proposed “pool[ing] contributions” from multiple sources to create standardized accounts that follow individuals across jobs and industries, ensuring continuity of healthcare coverage and retirement accumulation even as workers move between employers or between employment and non-employment. The company argued such pooled funding structures would better reflect the likely future of work in an AI-driven economy.
On regulation, OpenAI called for measures that protect children, guard against national-security threats and avoid entrenching incumbent firms. The company emphasized rules should be “common-sense” and not written in ways that effectively reward dominant market players, a nod to concerns that early regulatory frameworks can inadvertently raise barriers to competition.
OpenAI framed the proposals within a broader warning about the trajectory of AI. “We are entering a new phase of economic and social organization that will fundamentally reshape work, knowledge, and production,” the company wrote, urging governments, industry and civil society to “think boldly” and collaboratively develop an industrial policy agenda to ensure the benefits of superintelligence are broadly shared. The release came after CEO Sam Altman spoke at OpenAI’s Abilene, Texas, data center on Sept. 23, 2025, an appearance the company highlighted alongside the paper.
The proposals arrive against a backdrop of mounting labor anxiety and corporate restructuring in the tech sector. In recent weeks some large technology firms have announced job cuts tied to shifts toward AI investment, and polls show significant numbers of workers fear their jobs could be displaced by automation. OpenAI’s paper seeks to position the company as both a creator of transformative technology and a convener of policy debate about its societal consequences.
OpenAI’s suggestions will now enter a crowded and politically fraught field. Tax reforms, portable benefits and industrial policy are all contentious, and any serious movement would require legislative action and coordination across multiple levels of government. For now, OpenAI says the goal is to start the conversation early: to explore ideas and prompt debate on how to equip institutions for an AI-driven future.