Most commentary about Trump and Europe gets stuck on ideology—“the EU hates him” or “Europe is weak.” But this POLITICO report describes something more operational: allies treating perceived presidential unpredictability as a planning problem, not a political preference.

Here’s what happened: after a January 17 meeting at Mar-a-Lago, Slovak Prime Minister Robert Fico, one of the EU leaders most publicly friendly to Trump, privately relayed alarm to other European leaders. According to diplomats briefed on the exchange, Fico described Trump’s “psychological state” as concerning, using words like “dangerous,” “traumatized,” and “out of his mind.” The White House denies the characterization.

I’m not making a medical claim. I’m saying this: when counterparts, even sympathetic ones, come away describing volatility, uncertainty becomes the variable that drives alliance behavior. That’s the frame shift worth tracking: if the goal is stable deterrence and predictable trade, the focus moves to reliability of signals and institutions, not the tribal label of who’s “pro-Trump” or “anti-Trump.”

The machinery looks like this:

• Unpredictable threats → emergency summits and “reassess ties” declarations
• Credibility wobble → partners build alternatives, reduce dependence (defense investments, trade tools, contingency frameworks)
• Threats normalized → markets and governments start pricing in walk-backs… until the day there isn’t one

A key detail: this private exchange happened January 22, on the sidelines of an emergency EU summit convened after the Greenland/tariff crisis—a flare-up in which Trump remarked at Davos about using “excessive strength and force… but I won’t do that.” Reuters reports that even after the walk-back, EU leaders said the episode had shaken confidence. The Atlantic’s analysis tracks how the Greenland crisis specifically pushed European thinking toward tougher trade instruments and anti-coercion tools.

Question for SN: if allies believe U.S. unpredictability is rising, what’s the rational response—more deference to preserve access, or faster independence to reduce exposure?