Did Trump Use ChatGPT to Write the Tariff Bill?
Last week, Donald Trump unveiled a fresh wave of tariffs, and experts didn’t just raise eyebrows—they started running code. The tariff formula appeared suspiciously… robotic. As in, “Hey ChatGPT, how do I fix a trade deficit?” kind of robotic.
And guess what? When you plug the trade numbers into ChatGPT, Gemini, or Claude, you get a nearly identical outcome to Trump’s actual policy.
The Formula That’s Too Clean to Be Coincidence
Economist James Surowiecki tested it. He asked several leading AI models how to address trade deficits with specific countries. Their answer?
Take the trade deficit, divide it by the total imports from that country, then halve it. Boom: new tariff rate.
Sound familiar? That’s almost exactly how Trump’s new policy seems to be structured.
It’s the kind of logic only an AI—or someone heavily reliant on one—would propose: neat, linear, and completely disconnected from geopolitical reality.
So… Did He Actually Use ChatGPT?
No official confirmation, of course. But here’s what we know:
The tariff math mimics LLM-generated answers precisely.
The administration has given no coherent justification for the specific rates used.
The approach shows zero nuance around supply chains, trade retaliation, or global markets—all of which are LLM blind spots.
This doesn’t mean Trump or his team literally typed the bill into OpenAI and copy-pasted it. But someone very likely used an AI assistant for economic modeling—or worse, as a shortcut for building an ideological case. And it shows.
Why This Should Freak You Out
If AI tools are being used to build national economic policy without guardrails, we're facing a new class of political risk:
Over-simplified logic: LLMs aren’t trained in economic diplomacy. They don’t understand second- or third-order effects. Their suggestions look sound but often collapse under scrutiny.
Hallucinated accuracy: These models spit out answers with confidence, even when they’re wrong—like applying the same formula to Germany and Mexico without adjusting for their economic scale or trade relationships.
Zero accountability: If a policy causes chaos, no one can be held responsible for “asking the wrong chatbot.”
AI Policy is Coming. But Who’s Writing It?
This isn’t just about Trump. As LLMs like ChatGPT, Claude, and Grok infiltrate research, legislative drafting, and policy think tanks, we have to ask:
What happens when we automate authority?
Because if we let models trained on 2021 Reddit threads determine international tariffs, we might as well hand the nuclear codes to an AI trained on Marvel plot summaries.
🧠 Bottom Line: Whether Trump personally prompted ChatGPT or not, the fingerprints of AI-assisted thinking are all over this bill. The real concern isn’t just automation—it’s unquestioned automation. And that, ironically, might be the most human mistake of all.
Original video on tiktok by Nate.b.Jones
Subscribe and stay critical. Dream Big, Act Bigger.
Share this post