A New York-based firm wants to turn geopolitical chaos into something organizations can rehearse before it unfolds in the real world.
Anadyr Horizon has launched “Red Horizon,” a simulation capability that transforms geopolitical data into predictive, behavior-driven scenario models.
Built on the company’s North Star platform and powered by Agentic Systems Intelligence, the system models how real-world decisions made by governments, militaries, and markets could escalate into broader geopolitical crises.
With Red Horizon, users gain access to tailored simulations across political, economic, and security domains, alongside on-demand support to help interpret results and apply them to strategic decision-making.
Rather than analyzing events after they unfold, the platform examines how choices could evolve under pressure, revealing potential outcomes before they emerge.
According to the company, some users already employ Red Horizon to examine scenarios involving regional conflict escalation, financial exposure, and international security disruptions.
Forecasting Crisis Escalation
Red Horizon is said to have successfully modeled how the Iran war could escalate across military operations, intelligence activity, and foreign support networks.
The company also said the platform anticipated core dynamics of Russia’s invasion of Ukraine roughly 75 days before the conflict began, while separately flagging a maritime seizure scenario near Cuba days before it occurred.
“Most institutions don’t fail because they lack information. They fail because they misjudge how decisions will cascade under pressure,” said Dr. Arvid Bell, Anadyr Horizon Co-founder and Chief Executive Officer.
“Red Horizon gives them a way to see those cascades before they begin. It provides a persistent capability for navigating uncertainty.”
Harvard-Made Solution
The platform draws on more than a decade of research and simulation design work at Harvard University, where its underlying methodology was reportedly used to train senior military personnel, diplomats, and policymakers.
The simulations are based on research suggesting that modern crises are often driven less by static variables and more by chains of decisions that compound under pressure.
