Our suite of frontier AI models, trained on dozens of historical examples of institutional failure, reviews the most relevant events each day to warn you when to leave — before it’s too late.
Down 1 point from yesterday's 49.
Conditions deteriorating. Make your exit plan.
The Iran war enters week two with 18 U.S. dead, $37.5B spent, and now the House Speaker publicly breaking with Trump — institutional cracks are widening under wartime pressure.
When enforcement agencies fabricate evidence, wars expand unchecked, and political unity fractures simultaneously, the cost of waiting to make an exit plan rises faster than the cost of making one now.
Whether Congress moves to formally constrain the Iran operation or the administration escalates infrastructure strikes despite the Speaker's opposition — one stabilizes, the other accelerates the slide.
The current situation most closely mirrors Poland between 2018 and 2020, when the ruling party systematically degraded judicial independence and used state enforcement as a political weapon while opposition and courts still partially functi
The US sits in a mid-stage institutional capture phase — norms are actively eroding, enforcement is being weaponized, but elections remain competitive and some checks still produce real friction.
In Poland, this phase ended in one of two ways depending on electoral outcomes: either a democratic coalition eventually reversed course, or executive consolidation locked in structural changes that proved very difficult to undo.
YOUR PLAN
The score tells you the temperature. A plan tells you what to do about it.
Trump's Iran war has cost $37.5 billion with no end in sight
An open-ended war with mounting casualties and no defined exit conditions historically expands executive emergency powers at home while draining the political capital needed to check other abuses.
ICE agents fabricate assault charges and falsify evidence at alarming rates
When a federal enforcement agency routinely manufactures evidence against civilians and faces no meaningful accountability, it signals that the rule of law is no longer operating uniformly — a foundational condition for personal safety calculus.
House Speaker breaks with Trump, urges quick end to Iran war
A public split from within the president's own party is one of the few remaining institutional friction points, and whether it produces any actual constraint on war policy in the next 72 hours will reveal how much leverage Congress still holds.
FREQUENTLY ASKED
The Exit Signal Score is a daily 0–100 assessment of US stability generated by an ensemble of independent AI models. Each model reads the same daily briefing and scores the situation independently. The scores are averaged to produce a consensus, and all individual scores are displayed because no single model should be trusted alone.
The score is re-computed every day. A fresh daily briefing is generated from global news and social media signals, then sent to all multiple models for independent scoring. New scores typically appear each morning.
No. The Exit Signal Score describes present conditions, not a recommendation. It is a preparedness tool that tells you how worried to be — not what to do. Higher scores suggest more active preparation; they do not tell anyone to leave. CONTINGENCYPLAN.AI does not give financial, legal, immigration, or political advice.
Any single model carries its own biases, training cutoff, and blind spots. Running the same briefing through six independent models and averaging their scores reduces the influence of any one model’s idiosyncrasies. When all models agree, the signal is strong. When they disagree, that disagreement is itself useful information.
Every model receives the same daily briefing: clustered news events from global and US sources, summarized and categorized. Models do not have live internet access during scoring. They score based only on the briefing, so the score reflects the events in the briefing, not each model’s separate view of the world.