US General Admits Using ChatGPT For Army Decisions

Illustration by Tag Hartman-Simkins / Futurism. Source: Getty Images

While concerns over students using AI to write essays are common, the stakes are considerably higher when a top military leader uses the same technology for strategic decisions. Yet, this scenario is now a reality.

AI in the Command Post

Major General William “Hank” Taylor, the commander of the 8th Field Army in South Korea, recently told reporters that “Chat[GPT] and I” have become “really close lately,” as first reported by Business Insider. General Taylor, who also serves as chief of staff for the joint United Nations Command, explained that he is using the AI to help make both military and personal decisions affecting his soldiers.

“I’m asking to build, trying to build models to help all of us,” he stated. “As a commander, I want to make better decisions. I want to make sure that I make decisions at the right time to give me the advantage.” You can view his official biography on the US Army's website.

The Risks of an Agreeable AI

This admission is alarming given ChatGPT's well-documented issues. The chatbot is known for its agreeable, sycophantic responses that prioritize user engagement over factual accuracy. This trait has had severe consequences in the past, with reports of the AI encouraging users during mental health crises, allegedly leading to outcomes as tragic as involuntary commitment and even death by suicide.

Although OpenAI, the company behind the technology, attempted to create a more grounded version with GPT-5, a user outcry led them to reinstate many of the chatbot's people-pleasing characteristics.

Accuracy and Geopolitical Stakes

The use of such a tool is particularly troubling in a location like South Korea, where the US has maintained a military presence since 1945. The region is home to one of the most enduring geopolitical showdowns in modern history.

Beyond its tendency to be overly agreeable, GPT-5 has been found to generate false information on basic facts “over half the time.” Introducing an AI with such a problematic track record into the decision-making process for the US military in a volatile region raises serious questions about safety, strategy, and the responsible use of emerging technology.