ChatGPT's assessment of a world at risk

Details

In early May 2025, I engaged ChatGPT in a simple exercise: help assess the major geopolitical and economic risks that face us, their impact and their likelihood. In the conversation that ensued, ChatGPT suggested organizing the main points in a matrix, and after several iterations, with some of my suggestions incorporated, we came up with this:

Likelihood →
Impact ↓		 Very High High Moderate
Very High Global recession
(tariffs, fragmentation)		 Major international conflict
(China–Taiwan, India–Pakistan)		  
High U.S. domestic political crisis
(legal, electoral, unrest)		 EU crisis
(Far-right, debt, migration)		 NATO destabilization
Cyberattack
U.S.–Denmark
(over Greenland)
Moderate     Chinese pressure
(South China Sea)
Ukraine collapse
AI disinformation
Low     Iran–Israel war
U.S.–Canada
(economic conflict)
Legend
U.S.-centric Events primarily affecting or originating within the United States
(e.g., constitutional erosion, domestic unrest, U.S.–Canada friction).
EU/internal Western Threats rooted in or significantly affecting Europe or NATO's internal cohesion
(e.g., far-right politics, NATO weakening).
Regional (Geopolitical) Localized international threats that may spill over
(e.g., China–Taiwan, Iran–Israel, South China Sea disputes).
Global/Systemic System-wide threats affecting multiple regions or institutions simultaneously
(e.g., global recession, cyberwarfare, AI disinformation).
Shading Shading intensity reflects assessed severity:
darker = high/very high risk; lighter = moderate or secondary risk.

ChatGPT then went on to explain:

The structure of this matrix is not arbitrary. Each cell represents a convergence of two axes: the likelihood of a disruptive event occurring in the coming months, and the impact such an event would have if it did. What makes this particular visualization striking is its diagonal pattern: most of the events deemed high impact are also those considered highly likely.

This alignment suggests something unsettling about our current historical moment: the world is not bracing for a rare, unpredictable "black swan" event. Instead, it is navigating a landscape where systemic shocks are both visible and probable—where the biggest risks are not surprises, but expected outcomes of trends already in motion.

Among the most pressing concerns are:

  • A global economic downturn, driven by escalating tariffs, weakening global institutions, and deliberate fragmentation of trade alliances.
  • A U.S. domestic political crisis, where the convergence of legal defiance, electoral delegitimization, and civil unrest threatens the core stability of one of the world's oldest constitutional democracies.
  • The potential for major international conflicts, particularly in flashpoints like the Taiwan Strait or between India and Pakistan, where brinkmanship and strategic miscalculation could ignite large-scale war.
  • A widening European crisis, as far-right political movements, fiscal pressures, and fragmentation within NATO challenge the cohesion of the post-WWII European order.
  • And emerging global-systemic risks—from cyberwarfare to AI-driven disinformation—that erode trust, degrade coordination, and amplify instability.

But what gives this matrix its broader meaning is not just the identification of these threats. It is the pattern they reveal. Systems analysts and historians alike recognize this type of configuration: it resembles what physicists might call a system near a critical phase transition, or what political theorists might label a pre-crisis state. In such a system:

  • Small perturbations have disproportionately large effects.
  • Institutional buffers (like the rule of law, shared narratives, or alliance norms) are weakened.
  • Feedback loops—between economy, governance, technology, and security—begin to reinforce rather than regulate disruptions.

In short, this is what a world on the verge of nonlinear transformation looks like.

This pattern is not without precedent. History offers chilling analogues of periods when multiple stressors aligned across political, economic, and military domains. In July 1914, a tightly interconnected system of alliances and nationalist pressures, once resilient, unraveled in a matter of days—despite widespread belief that war was irrational. In the late 1920s, global markets exhibited high performance and high fragility, setting the stage for collapse. The fall of the Roman Republic similarly began with institutional rot and polarization, long before any single coup or civil war. In each case, the warning signs were not hidden—they were overlooked, normalized, or dismissed.

To be clear, none of this is deterministic. Societies can and have pulled back from the brink. But the risks enumerated here are not abstract probabilities—they are the shape of trajectories already underway. What remains unknown is whether the institutions and actors involved will course-correct—or accelerate.

This matrix, then, is not merely a diagnostic tool. It is a snapshot of a system under stress, a moment in which understanding the structure of risk is the first step toward mitigating it.