Decision Tree Research is an independent research and analysis firm that specializes in strategic technologies, public policy, and geopolitics.
“Allen has been a must-read.”
“Very much at the center of the formulation of current U.S. policy.”
“The most-watched U.S. analyst on the subject.”
Subscribe to Gregory C. Allen’s newsletter for timely, evidence-based analysis of the big policy debates in AI, semiconductors, and defense tech.
Our areas of focus
AI capability has become a measure of national power and a geopolitical race for superiority. The competition covers frontier model development, U.S.–China rivalry, industrial policy, export controls, military integration, and the international safety and governance architecture forming around them.
Chips have become the new oil. The contest for silicon supply chain security spans advanced chips, the equipment and materials that produce them, and the export controls and industrial policy reshaping the global industry.
The defense industrial base is being rewritten by a new generation of commercial AI and space firms. Military AI, autonomous weapons, drones, and the commercial space revolution are reshaping the strategic balance.
Featured research
How Big Is China’s Reported DUV Semiconductor Lithography Breakthrough?
China’s first domestically produced immersion DUV lithography tools are real — but the breakthrough is smaller than the drop in ASML’s share price implied, and bigger than China’s official silence suggests.
Four keys to understanding the future of AI and unemployment
A data-driven crash course in the most illuminating historical analogies for AI and employment — from farm mechanization and the fate of the horse to the economics of music and chess.
How Much Did U.S. Semiconductor Export Controls Accelerate China’s Push for Self-Sufficiency?
In terms of Chinese government subsidies, not at all: investment data shows a semiconductor self-sufficiency push that was already running at full speed before U.S. export controls arrived.
From The AI Power Podcast
Weekly episodes on AI policy, semiconductors, and the U.S.–China technology competition — hosted by Gregory C. Allen.
Nvidia CEO Hangs Out with Chinese Spy? + China’s New AI Export Controls + The AI/Cyber Month from Hell + AI in the Midterm Elections
Greg and Adam Goodwin dig into the New York Times report that Chinese intelligence targeted Jensen Huang through the CEO of an Nvidia partner company, and the long list of red flags around efforts to get Nvidia chips into China. Then they break down China’s new AI export controls, which bear a striking resemblance to the Biden-era Diffusion Rule, and what Beijing wants from them. Greg also recounts the PHP vulnerability and a record-setting “Black October” for AI-found vulnerabilities, and the episode closes on how voters feel about AI heading into the midterms.
Listen →How China Got 340 DUVi Lithography Machines and Why We Should Stop Selling Them with Saif Khan and Nick Brown
Everyone talks about EUV, but China’s AI chips are made on older immersion DUV machines, and it now has around 340 of them. Saif Khan and Nick Brown of the Center for Technology and Statecraft join Greg to break down their 100+ page report on how China built that stockpile and why they say a countrywide export ban is the fix. They also cover whether China can build its own, and what it would take.
Listen →Trump Likes Mandatory AI Regulation for America but Not China? + Anthropic Defeats Distillation
Greg and Jack Conlin break down how an incident involving Anthropic’s AI agents pushed the White House’s Super Intelligence Task Force toward tougher de facto regulation and oversight of AI, and what CAISSI’s new name and more direct line to the White House mean. Greg argues that tightening the rules on U.S. models while leaving Chinese models unaddressed is inconsistent and puts American AI leadership at risk. Then: Anthropic reports it has stopped 90–99% of adversarial distillation, a core avenue for Chinese AI development. Will that advantage last?
Listen →Why “Decision Tree”?
The decision tree is one of the most useful analytical tools in three disciplines that rarely meet: Machine Learning & AI, Corporate Strategy, and Government Policy Analysis.
In machine learning, the decision tree is foundational to how models classify and predict. In business strategy, it structures operations planning, capital allocation, and risk decisions. In policy analysis, it maps the consequences of regulation, sanctions, and other interventions through complex systems.
Gregory C. Allen’s career has spanned all three. The firm’s name signals the methodology that runs through the work: enumerate the decisions, trace the branches honestly, find the paths that carry the most consequence, and reason rigorously before acting. Decision Tree Research applies that discipline to our analysis of the technologies shaping global power competition.
Built on a body of work
Before founding Decision Tree Research, Gregory C. Allen authored some of the most impactful analysis in U.S. AI policy — at CSIS, the Department of Defense Joint Artificial Intelligence Center, the Center for a New American Security, and the Belfer Center for Science and International Affairs at Harvard Kennedy School.