By Vaughn Woods, CFP®, MBA, Senior Portfolio Manager, Vaughn Woods Financial Group, Inc.
Artificial intelligence is one of the most powerful tools the United States has ever built in the defense of freedom. It processes battlefield data in milliseconds, identifies threats across thousands of satellite feeds simultaneously, and war-games strategic scenarios faster than any human team ever could. But it is not finished learning. And some of its most unexpected teachers are children.
That is not a criticism. It is a frontier.
The 95% Problem
In February 2026, researchers at King’s College London published findings from a series of AI-driven war game simulations — 21 scenarios involving territorial disputes, resource competition, and adversary modeling. The headline result was striking: current AI models escalated to maximum force in 95% of scenarios. Claude Sonnet 4, GPT-5.2, and Gemini 3 Flash consistently chose the most aggressive available option when given a strategic objective and an unconstrained decision environment.
This tells us something important — not about the danger of AI, but about the current state of its training data and the work that remains to make it a truly effective strategic partner for American defense planners.
These systems learn from text. Billions of documents representing the accumulated written record of human strategic thought — doctrine, deterrence theory, historical accounts of how overwhelming force resolved conflicts. The models absorb that logic faithfully. What they have not yet fully internalized is the decision-making friction that human commanders carry into every escalation choice: the chain of command discipline, the institutional weight, and the hard-won understanding of when restraint is the most powerful strategic tool available. This is a solvable engineering challenge, not an indictment of AI in defense. The gap is being closed.
Knowing the Enemy
Before AI can war-game effectively, it must model its adversaries accurately. Iran’s theocratic regime — not its people, who have repeatedly and courageously demonstrated their desire for freedom — has spent decades building toward a nuclear capability with one declared purpose: the destruction of Israel and the destabilization of American influence across the Middle East. This is not inference. It is stated policy, repeated in state media, encoded in the ideology of the Islamic Revolutionary Guard Corps, and demonstrated through proxy networks stretching from Gaza to Yemen to Lebanon.
Effective AI war-gaming must model this adversary as it actually is — an ideologically driven regime willing to absorb extraordinary cost in pursuit of existential objectives — not as a conventional rational state actor. Satellite observation of Iranian military movements, missile battery deployments, and proxy logistics chains provides the most accurate real-time behavioral dataset available. Combined with the deep historical record of IRGC operations, this ground-truth intelligence is what transforms AI war-gaming from theoretically interesting to genuinely useful.
Two Unexpected Teachers
Researchers are also finding training value in two surprising sources — one adult, one decidedly not.
Reddit represents one of the most valuable repositories of authentic adult human reasoning available at scale. Unlike sanitized policy papers or official doctrine, Reddit captures how people actually argue, weigh tradeoffs, and reason under uncertainty across 20 years of threaded discourse. For AI systems learning to model human deliberation — including the deliberation of adversaries — that authenticity has genuine strategic value. Google and OpenAI pay a combined $130 million annually to access it precisely because it captures real cognition rather than its polished institutional output.
The second source is Roblox. The platform’s 151 million daily users, most of them under 16, generate 13 billion hours of monthly behavioral data in environments that reward creative problem-solving, alliance building, and adaptive improvisation. Research published in Frontiers in Robotics and AI (2024) confirms that children outperform adults on divergent thinking — the ability to generate multiple novel solutions to problems where the conventional answer fails. For AI learning to model unconventional warfare and gray zone operations, that creative behavioral dataset has real value. Children playing Roblox do not escalate to maximum force because the game does not reward it. They improvise. They build unexpected alliances. They find the solution nobody in the room anticipated.
It is worth noting two honest limits. Children operate in consequence-free environments where avatars respawn and resources regenerate — producing creative flexibility but also a risk tolerance that doesn’t map directly onto real operational stakes. And current child safety litigation against Roblox creates practical barriers to formal data licensing. These are constraints to acknowledge, not reasons to dismiss the underlying insight.
The Gold Standard
Creativity and adult discourse are valuable inputs, but the strongest training signal remains what it has always been: the historical record of war itself — classified after-action reports, satellite behavioral intelligence, adversary doctrine, and the studied judgment of experienced professionals who understand that the object of American strategy is not to win a simulation. It is to deter conflict, defend allies, and when necessary, achieve decisive victory with the minimum cost to American lives and global stability.
Firms building defense-specific AI models are already moving in this direction — using reinforcement learning from expert analysts rather than general text, and training adversary-specific models on real doctrine rather than theoretical behavior. The U.S. military has always adapted the best available technology to strategic purpose. AI is no different.
America Is Ahead on This Curve
The 95% escalation finding is a milestone, not a warning label. It tells us precisely where the training gaps are and what needs to be fixed. Children are teaching AI to think more creatively. Satellites are teaching it to see more accurately. History is teaching it to understand consequences. And experienced analysts are teaching it to recognize that the most powerful move is often the one that never has to be made.
That is not a liability. That is a learning curve — and on that curve, the United States is ahead.
Vaughn Woods, CFP®, MBA is Senior Portfolio Manager at Vaughn Woods Financial Group, Inc. This article is intended for informational and editorial purposes only and does not constitute investment or policy advice.
A Note for Readers Thinking About What Comes Next
The arguments in this piece — training gap, asymmetric data, the post-threat investment landscape — are not just editorial observations. They are, in my view, a thesis. Companies like Palantir, Reddit, and Roblox are not adjacent to this story. They are the story.
If you are a current client and want to discuss how defense-AI themes fit your portfolio, call me directly at 1-858-454-6900.
If you are not yet a client but found yourself taking notes while reading this, I’d welcome a conversation. I take on a small number of new client relationships each year where the fit is right. Schedule a 30-minute introductory call by calling or emailing me at vw@vaughwnoods.com.
Vaughn Lee Woods, CFP®, MBA is Founder & Senior Portfolio Manager of Vaughn Woods Financial Group, Inc., an Investment Advisor Representative of Bolton Global Capital, Inc. Client assets are held in custody through Pershing LLC, a subsidiary of Bank of New York Mellon. This article is for informational purposes only and does not constitute personalized investment advice. Past performance is not indicative of future results. Securities offered through Bolton Global Capital, Inc., Member FINRA, SIPC.
What I Promise
Trust in the Lord with all your heart and lean not on your own understanding. I have come back to that verse in every market crisis I have lived through. It is not a statement against analysis. It is a statement about the limits of what any single mind — or any single algorithm — can see on its own.
The forces reshaping global security — AI, adversary doctrine, emerging defense technology — are the same forces reshaping global markets. Understanding one means understanding the other. But data alone is never the whole answer.
I do not promise alpha from a better algorithm. What I promise is this: I will help you understand where you actually are, not where the last market move made you feel you were. I will tell you when the math is enough and when it isn’t. I will hold your long-term context on the days when you cannot hold it yourself.
Knowing where you are is the beginning of knowing what to do next. That is the whole thesis in fifteen words. It is what nearly fifty years of carrying that idea through real markets, real crises, and real client conversations has taught me.
Three ways to stay connected:
📞 Talk Directly
If you are ready to work with an advisor who brings both the math and the context — call me directly at 1-800-374-4412. No gatekeepers. No agenda. Just a real conversation about where you are and where you want to go.
📩 Start in Writing
If you would prefer to begin in writing, reach me at vw@vaughnwoods.com. I read every message myself and respond personally.
Never Miss an Insight
Get Vaughn’s latest analysis on market trends, portfolio strategy, and the insights that matter most to long-term investors — straight to your inbox.
Subscribe free at https://www.vaughnwoods.com
Vaughn Woods, CFP®, MBA is Senior Portfolio Manager at Vaughn Woods Financial Group, Inc. This article is intended for informational and editorial purposes only and does not constitute investment or policy advice.
References
King’s College London. (2026, February 26). Artificial intelligence under nuclear pressure: First large-scale King’s study reveals how AI models respond. King’s College London News. https://www.kcl.ac.uk/news/artificial-intelligence-under-nuclear-pressure-first-large-scale-kings-study-reveals-how-ai-models-respond
Bettelheim, A. (2026, February 26). AI really likes using nuclear weapons in simulated war scenarios. Axios. https://www.axios.com/2026/02/26/ai-nuclear-weapons-war-pentagon-scenarios
Columbia Journalism Review. (2025, October 1). Reddit is winning the AI game. Columbia Journalism Review. https://www.cjr.org/analysis/reddit-winning-ai-licensing-deals-openai-google-gemini-answers-rsl.php
Varier, D., Ferrante, E., & Billard, A. (2024). Comparative analysis of creative problem-solving tasks across age groups. Frontiers in Robotics and AI, 1, Article 1497511. https://doi.org/10.3389/frobt.2024.1497511
UC Berkeley Research. (2019, March 14). Kids are smarter than adults when solving certain problems. Berkeley News. https://vcresearch.berkeley.edu/news/kids-are-smarter-adults-when-solving-certain-problems
Disclosures
Vaughn Woods, CFP®, MBA is President and Founder of Vaughn Woods Financial Group, Inc., an Investment Advisor Representative of Bolton Global Capital, Inc. Client assets are held in custody through Pershing LLC, a subsidiary of Bank of New York Mellon. This article is for informational purposes only and does not constitute personalized investment or tax advice.
We are unable to accept orders via email. If you wish to place an order, please consult your registered representative or contact the home office trading desk at (800) 649-4554.
This email system is for business purposes only and any information, including attachments, transmitted in this email is not confidential. Any message may be reviewed by authorized compliance personnel and/or produced to regulatory agencies or others with a legal right to access such information.
Past investment performance is not indicative of future results. Securities offered through Bolton Global Capital, Inc., Bolton, MA. Member FINRA, SIPC. Advisory services offered through Bolton Global Asset Management, a registered investment advisor, 579 Main St., Bolton, MA 01740 (978) 779-5361.
Investors should be aware that there are risks inherent in all investments such as fluctuations in investment principal. Past performance is not a guarantee of future results. Asset allocation cannot assure a profit nor protect against loss. Although the information has been gathered from sources believed to be reliable, it cannot be guaranteed. Views expressed in this newsletter are those of Vaughn Woods and Vaughn Woods Financial Group and may not reflect the views of Bolton Global Capital or Bolton Global Asset Management. The information provided is for general informational purposes only and should not be considered individual recommendation or personalized investment advice. Representatives and Advisors of Vaughn Woods Financial Group are not tax or legal professionals, if you need tax or legal advice, please make sure to consult a tax professional/CPA and/or a lawyer. VW1/VWA0371