AI risks in warfare demand new global rules

Former Foreign Minister Yun Byung-se, left, speaks during a session on AI's role in the military domain at the Doha Forum 2025 held in Doha, Qatar, Saturday. From right are professor Nehal Bhuta at the University of Edinburgh; professor Jeroen van den Hoven at the Delft University of Technology; professor Michael C. Horowitz at the University of Pennsylvania; and professor Denise Garcia at the Northeastern University. Korea Times photo by Baek Byung-yeul

DOHA — As artificial intelligence (AI) rapidly reshapes the landscape of modern warfare, experts at the Doha Forum 2025 held Saturday called for immediate international action to establish rules before AI destabilizes global security.

Their discussion took place during a session themed "AI as a Double-Edged Sword: Strategies for Responsible Use in the Military Domain,” which was held in partnership with the Jeju Peace Institute, a leading Korean think tank. The panelists explored the ethical, legal and security implications of integrating AI into military systems.

Yun Byung-se, former foreign minister of Korea and current chair of the Global Commission on Responsible AI in the Military Domain (REAIM), said that AI is a transformative force comparable to the discovery of fire, not just a simple tool.

"AI is no longer a niche technology. It is a civilizational technology, reshaping nearly every domain of human activity like an unbound Prometheus," Yun said. "As a dual-use, transformative technology, AI offers extraordinary opportunities, while at the same time introducing deeply consequential risks."

The former minister stressed that using AI in military decisions is much more dangerous than using it in civilian life because it poses a particular danger with regard to nuclear weapons systems.

"Its possible integration with decision-making on nuclear weapons evokes deep concerns about humanity's future," he said. "This raises fundamental questions of accountability, oversight and human control.”

Yun introduced the commission’s recent report "Responsible by Design," which aims to provide a roadmap for governance. He said that governing military AI is difficult and lagging behind civilian efforts, but cited recent progress, such as United Nations resolutions supported by more than 160 countries.

"Humanity must learn to govern AI before AI governs humanity," Yun emphasized. "The future trajectory of our civilization will be determined by the choices we make now."

From left are former Foreign Minister Yun Byung-se, professor Michael C. Horowitz, professor Denise Garcia, professor Nehal Bhuta and professor Jeroen van den Hoven after participating in a session at the Doha Forum 2025 held in Doha, Qatar, Saturday. Korea Times photo by Baek Byung-yeul

The talks shifted to the nature of the technology. Michael C. Horowitz, a professor at the University of Pennsylvania, argued that military AI is a general-purpose technology, like electricity or the combustion engine, and not just one type of weapon.

"It’s not a widget, it’s not a weapon. It’s a general-purpose technology," Horowitz said. "Nobody talked about an electricity arms race ... so we shouldn't expect this to be different in the AI context."

Addressing the ethical dimension, professor Jeroen van den Hoven at the Delft University of Technology said technology cannot be held responsible for moral choices.

"Responsible AI is a form or type of AI that supports human beings as responsible agents, and it is designed to exactly do that," van den Hoven said. "It needs to be designed for responsibility."

Nehal Bhuta, the chair of public international law at the University of Edinburgh, gave a compelling legal viewpoint that acknowledged both the difficulties and hope regarding future AI regulation.

"I’m pessimistic about the prospects of new legal frameworks. it’s difficult to get states to converge around new norms," Bhuta explained.

The Korea Times is participating in the forum as a media partner.