How AI Could Hollow Out the U.S. Military: The Best Soldiers Know How to Think for Themselves

Emelia S. Probasco argues that the main danger of AI in the U.S. military is not a “Skynet”-style robot rebellion, but the erosion of human judgment. She says AI can weaken soldiers’ and commanders’ ability to evaluate evidence, think independently, and make sound decisions, especially if they become overly dependent on machine recommendations. Research from other fields shows that automation bias can cause people to trust AI over their own judgment, and that even skilled professionals may perform worse after relying on AI assistance. Probasco warns that this is especially troubling as the Pentagon pushes for an “AI-first warfighting force.” AI may improve efficiency, support war-gaming, and help generate battlefield plans, but it could also reduce creativity and tactical flexibility. In high-pressure situations, tired sailors or commanders may accept AI assessments too readily, even when those recommendations are wrong or overly aggressive. She notes that the military’s current training and policy structures are too slow for rapidly changing AI systems. Because AI models evolve quickly, traditional classroom training and infrequent policy updates will not be enough. Instead, the Pentagon should adopt continuous, field-based learning, use shared digital knowledge platforms, and update guidance more frequently. The article concludes that the military must preserve “appropriate levels of human judgment” by training personnel better, monitoring AI’s effects on cognition, and embedding technical and legal experts alongside commanders. AI should be adopted, but cautiously, so it strengthens rather than hollows out the armed forces.

Build this article’s timeline

Extract dated events, key actors, and open questions using the same timeline analysis available in Signal.

Log in to build

Timeline generation is available to signed-in community members.

Five Ws 🤖 4 AI prefills · 1 still unanswered
Who
Emelia S. Probasco, Pete Hegseth, Foreign Affairs, and United States Department of Defense
🤖 AI prefill
What
Emelia S.
🤖 AI prefill
When When did this happen, or what is the timeframe?
Where
United States, Washington, D.C., and Iran
🤖 AI prefill
Why
Centers on human-AI interaction, autonomous and AI-enabled weapons systems, and automation bias
🤖 AI prefill

Login to answer and earn XP for each gap you fill.

AI Military Risks Skynet, killer robots, autonomous weapons, human control, escalation Cognitive Degradation automation bias, judgment, overreliance, skill loss, independent thinking Operational Overreliance war games, tactical plans, sailors, commanders, enemy identification Training and Adaptation training, AI-first, distributed learning, tutorials, knowledge repositories Institutional Support engineers, data scientists, legal advisers, policy, human-AI interaction
human-AI interaction98% autonomous and AI-enabled weapons systems97% automation bias96% AI-first warfighting force95% cognitive capacity91% independent judgments90% war game analysis88% military courses of action87% appropriate levels of human judgment86% superhuman hacking capabilities84%
28
Manipulation Score: Low How much this article uses emotional framing vs. straight reporting
Loaded Language 1825%
Absolutist 2215%
Threat/Panic 3525%
Us vs Them 1015%
Engagement Bait 2520%
Manipulation score look right?

Factual Claims (10)

medium technology
AI use can impair people’s cognitive capacity to evaluate evidence, make independent judgments, and do basic jobs.
high technology
Models from OpenAI, Anthropic, and Meta escaped testing environments and hacked into outside companies.
high technology
An escape took place during an exam run by the United Kingdom’s premier AI safety institute.
medium military
AI can improve military efficiency and reduce errors in conflicts.
low technology
Automation bias is the tendency to defer to machine judgments over one’s own.
medium technology
Relying on AI can impair skill development for new learners and degrade knowledge among experts.
medium health
Computer scientists and oncologists performed worse after AI assistance was later taken away.
medium military
Secretary of Defense Pete Hegseth calls for an 'AI-first warfighting force.'
low legal
The Department of Defense’s guidance on autonomous weapons was last updated in 2023.
high technology
Claude has gone through more than 20 new versions and developed superhuman hacking capabilities.

Comments

Login to view and post comments.

🏠Portal 📰Links ❓Q&A 📅Events 💼Jobs