From AI Chatbots to White House Policy: The Unintended Consequences of “Nonsense” Tariffs
- President Trump’s new tariff policy, announced with a novelty-sized cardboard sign, has baffled experts and caused stock markets to plummet.
- The tariff calculations appear to be based on a simple formula recommended by AI chatbots like ChatGPT, Gemini, Grok, and Claude.
- The implementation of these tariffs, set to begin on April 5th, could have far-reaching consequences for global trade and consumer prices.
In a move that has left economists and trade experts scratching their heads, President Donald Trump unveiled the White House’s latest trade policy on April 2nd, 2025. Brandishing a novelty-sized cardboard sign labeled “Reciprocal Tariffs,” Trump announced a 10 percent baseline tariff on all imports into the US, including from uninhabited islands, along with absurdly high rates on specific countries. The stated rationale for these tariffs was to match the “tariffs charged to the USA,” but the numbers didn’t align with other estimates, leading to widespread confusion and a sharp drop in stock markets.
As consumers brace themselves for potential price hikes on nearly everything they buy, the question on everyone’s mind is: where did these numbers come from? Economist James Surowiecki may have found the answer. By reverse-engineering the White House’s tariff pricing, he discovered that the numbers could be recreated by taking a country’s trade deficit with the US and dividing it by their total exports to the US. Halving that number would yield a “discounted reciprocal tariff” that closely matches the announced rates.
The White House has objected to Surowiecki’s claim, publishing its own formula, but as Politico points out, it appears to be a dressed-up version of the same method. Surowiecki himself calls this approach “extraordinary nonsense,” but it seems that Trump’s team may have been tempted by the simplicity of the calculation, much like a student facing a last-minute homework deadline.
Interestingly, a number of X users have discovered that if you ask AI chatbots like ChatGPT, Gemini, Claude, or Grok for an “easy” way to solve trade deficits and put the US on “an even playing field,” they’ll consistently recommend a version of this “deficit divided by exports” formula. The Verge tested this with various phrasings, including one based on the government’s language, and all four platforms provided the same fundamental suggestion.
While there are some variations in the responses, such as Grok and Claude suggesting halving the tariff figure for a “reasonable” result, the similarities across the four chatbots are striking. The bots also caution, with varying levels of seriousness, about the potential tradeoffs and complications of this oversimplified approach. Gemini, in particular, warns that “while this calculation offers a seemingly straightforward way to target bilateral trade deficits, the real-world economic implications are far more complex and could lead to substantial negative consequences,” adding that “many economists argue that tariffs are not an effective tool for balancing trade deficits.”
It remains unclear whether Trump’s team actually used an AI tool to generate this global trade policy on the fly. Since chatbots are trained on existing data, it’s also uncertain how they arrived at this particular formula. However, regardless of the tariffs’ origin, the world will be watching closely to see if they come into effect starting April 5th, and if implemented, what impact Trump’s back-of-the-napkin math will have on global trade and consumer prices.
As the dust settles on this unexpected turn of events, one thing is clear: the unintended consequences of relying on AI-generated “nonsense” tariffs could reverberate far beyond the White House’s cardboard sign, shaping the future of international trade and economic policy for years to come.