OpenAI warns that the transition toward superintelligence will come with serious risks

OpenAI is calling for taxes on robots and a series of policy changes to prepare humans for superintelligence and the disruptions that may come with it. 

In a publication Monday, OpenAI affirmed that the transition from frontier AI models to superintelligent ones has already begun. These are models capable of “outperforming the smartest humans even when they are assisted by AI,” it said. 

The company itself, which is also working on building superintelligent models, fears that such a technology comes with serious risks, from job losses to economic disruptions and cybersecurity misuse. “Without effective mitigation, people will be harmed,” it wrote, calling for safeguards and governance frameworks.

OpenAI proposals include taxing robots

OpenAI proposed several policy changes, including taxing robots. It precisely called for taxes related to automated labor, amid the growing fear of AI and robots replacing human laborers in offices and factories. 

The company also proposed shifting the tax base away from labor income and payroll. As robots and AI models take over production, OpenAI believes there will be an expansion of capital gains and corporate profits. In this light, it said policymakers should instead rebalance the tax base by increasing taxes on capital gains and corporate income.

OpenAI also suggested that policymakers and AI companies set up a so-called Public Wealth Fund that invests in AI companies and other firms that are adopting and deploying AI, as a way to give every citizen exposure to the growth of the AI economy. It said proceeds from the Funds should be distributed directly to all the citizens. 

As workload declines and operating costs fall due to AI use, OpenAI said the government should encourage companies to start trying a four-day workweek, with no loss in pay for workers. It also said companies should consider offering predictable “benefits bonuses” to encourage productivity improvements with AI tools among employees.

OpenAI may sound concerned about the adverse outcome of superintelligent models, but the firm itself is racing towards building one. In a blog post in December, Sam Altman mentioned that they will certainly have achieved superintelligence within the next 10 years. 

“Ten years into OpenAI, we have an AI that can do better than most of our smartest people at our most difficult intellectual competitions,” Altman wrote. “In ten more years, I believe we are almost certain to build superintelligence.”

White House releases its AI legislative framework

OpenAI’s policy recommendations come as U.S. policymakers are preparing to debate the AI Framework, a preemptive legislative recommendation that will govern AI developments and uses across the country. 

Many states in the U.S. have gone on to make their own AI laws in the absence of a uniform national approach. In 2025, however, President Donald Trump issued an  Executive Order (EO), directing the establishment of a national standard that would preempt existing state AI laws deemed burdensome to U.S. dominance in the global AI race. 

On March 2oth, the White House released a comprehensive national legislative framework to patch the conflicting state laws. The framework covers so many key areas, would be reviewed by several Congressional committees before it can be implemented. 

Your keys, your card. Spend without giving up custody and earn 8%+ yield on your balance with Ether.fi Cash.