Global Action Logs
Global Action Logs is a dataset that records the daily actions of 11 leaders across 12 countries.
Various actions—such as statements, visits, and agreements—are organized in chronological order.
Each record is structured around core fields including subject, action (verb), keywords, region, counterparty, and date.
By focusing strictly on “who did what,” and excluding interpretation or analysis, the dataset allows users to track flows,
changes, and differences in behavior over time.
The data is also optimized for AI-based analysis, enabling use cases such as trend identification, profiling, and cross-thematic analysis.
Format: JSON
Donald Trump
Xi Jinping
Vladimir Putin
Sanae Takaichi
Keir Starmer
Emmanuel Macron
Narendra Modi
Mohammed bin Salman
Benjamin Netanyahu
Vlodymyr Zelensky
Kim Jong Un
United States
China
Russia
Japan
North Korea
Israel
Iran
Ukraine
India
United Kingdom
France
Saudi Arabia
Actual log data (sample)
This example demonstrates how Global Action Logs can be used to generate a monthly behavioral profile using AI.
The snapshot below is based on the January 2026 log for Donald Trump, combining a fact-based summary with counterparty frequency analysis.
Identification of key actions and patterns
Frequency analysis of counterparties (countries, organizations)
Extraction of dominant verbs and themes
By structuring actions in this way, the dataset enables clear tracking of behavioral patterns and relationships over time.
This example demonstrates how Global Action Logs can be used as structured input for AI-driven scenario analysis.
Using action logs covering Nov 2025 to Feb 2026 for Israel, Iran, and the United States, an AI model was able to:
Extract key behavioral patterns from observed actions
Identify recurring dynamics across multiple actors
Generate plausible strategic scenarios based on these patterns
Rather than relying on narratives, this approach starts from structured records of actions and builds analysis directly from observed behavior.
This is not a prediction, but an example of how structured action data can be used to generate scenario-based insights.
This example demonstrates how Global Action Logs can be used to identify shared themes across multiple countries using AI.
The visualization below is based on the January 2026 logs for Israel, the United States, and Iran.
Extraction of recurring cross-country themes
Identification of common patterns (e.g., sanctions, military actions, protests)
Structuring and comparison of actions across different actors
By organizing actions into structured records, the dataset enables cross-country comparisons and thematic analysis at scale.
Beyond this use case, the dataset can also be used for:
Tracking changes in a single leader’s behavior over time
Identifying shifts in geopolitical focus (e.g., regions or topics)
Detecting emerging patterns across events and actors
Supporting AI-driven profiling and scenario generation
Generating structured inputs for storytelling and narrative-based content (e.g., video scripts)
Download a small sample dataset :
Latest updates and use cases are shared on X