In April 2025, the Trump administration introduced a sweeping set of tariffs, including a universal 10% import duty and significantly higher rates for specific countries. These tariffs, termed "reciprocal," were presented as a means to balance trade deficits. However, the methodology behind these tariffs has sparked controversy, with speculation that the formula used was generated by artificial intelligence tools like ChatGPT.
The Tariff Formula and AI Similarities
The controversy began when analysts noticed that the tariff rates appeared to follow a simplistic formula: taking a country's trade deficit with the U.S., dividing it by that country's exports to the U.S., and halving the result to determine the tariff percentage. This method closely resembled outputs from AI models when prompted to design reciprocal tariff structures.
As The Verge reported, tools like ChatGPT, Claude, Gemini, and Grok generated strikingly similar formulas when asked to create “fair” tariffs. Economist James Surowiecki criticized the logic as “extraordinary nonsense,” likening it to something an AI might hallucinate.
While the White House denied using AI in crafting the policy, the resemblance fueled speculation that generative AI tools had indirectly influenced the administration’s thinking.
Economists' Concerns and Misapplication of Research
Further scrutiny revealed that the administration may have misapplied academic research to justify the tariffs. Brent Neiman, an economist at the University of Chicago, said that his work had been misinterpreted by the Trump team, leading to an inflated and inaccurate formula.
A deeper critique was provided by the American Enterprise Institute, which argued that the formula ignores the legitimate economic reasons behind trade deficits and misrepresents basic principles of global trade.
Implications for Policy and AI Integration
Whether or not AI directly created the tariff formula, this controversy raises serious questions about the role of generative AI in policymaking. While tools like ChatGPT can rapidly generate plausible-sounding ideas, they lack deep economic understanding and are prone to “hallucinating” facts or oversimplifying complex systems.
As The Guardian pointed out, policy made by bots risks sounding rational but failing catastrophically in real-world application. The lack of transparency in how AI is used in government settings makes this even more concerning.
Conclusion
The AI-generated tariff formula may not have been official, but its resemblance to Trump’s trade policy is hard to ignore. This incident highlights both the power and the risks of integrating generative AI into political and economic decision-making.
As AI continues to evolve, it’s crucial that humans—not algorithms—remain firmly in charge of interpreting, contextualizing, and applying complex policies.