The Bulletin warns of global risks after Miwon Peace Prize win

The Bulletin of the Atomic Scientists, the organization behind the Doomsday Clock, received this year’s Miwon Peace Prize on Monday, warning that nuclear threats, pandemics, climate change and artificial intelligence (AI) are deepening interconnected risks to humanity.

The award ceremony was held at Kyung Hee University in Seoul during the 2026 Peace BAR Festival, the 45th international conference commemorating the U.N. International Day of Peace. The event was hosted by Kyung Hee University System.

Accepting the prize on behalf of the Chicago-based nonprofit, Alexandra Bell, its president and CEO, said her organization shared the late Kyung Hee founder Choue Young-seek’s view that humanity must move beyond narrow national interests and build a shared commitment to peace.

“The dangers created by modern science and warfare cannot be solved solely by nation-states … Their growing complexity requires international cooperation, the participation of all levels of society and an understanding that the stakes are no less than our shared survival,” Bell said.

The Miwon Peace Prize was established in 2024 to honor Choue’s philosophy and practice of peace. It is awarded biennially to an individual or organization judged to have contributed to world peace and the advancement of civilization. The inaugural recipient was The Elders, a group of independent global leaders founded by Nelson Mandela.

Founded in 1945 by scientists associated with the Manhattan Project, the Bulletin is best known for its Doomsday Clock — created in 1947 as a visual symbol of humanity’s proximity to self-inflicted catastrophe. The Clock is currently set at 85 seconds to midnight, its closest-ever setting.

While it does not predict the future, Bell said, it is an assessment of present conditions and a warning that action remains possible.

“Midnight means there is nothing left to do, nothing left to try to save humanity,” she said. “But I also see the time left on the Clock as an opportunity to change things. The time on the Clock is a lifeline, a ladder, a way out.”

Bell said the Bulletin’s concerns about humanity’s own destruction now extend well beyond nuclear weapons.

“After generations of struggle to reduce nuclear risks, those risks are on the rise, and a multipolar nuclear arms race that was not on the horizon is right here at our door. The quickening pace and devastation of extreme weather caused by climate change can be felt every day, everywhere around the world,” Bell said.

“Emerging and disruptive technologies hold the promise of a brighter future, but just like we saw at the dawn of the atomic age, the interest in advancing these technologies is progressing faster than our means to understand and control their potentially ill effects.”

In a congratulatory video message, former U.N. Secretary-General Ban Ki-moon praised the Bulletin’s work and said the organization’s annual Doomsday Clock had become “a global benchmark for existential risk.”

Ban, who accepted the inaugural Miwon Peace Prize on behalf of The Elders two years ago, said the current state of the Clock suggests a failure of leadership.

“The people of the world deserve a better future from their heads of state and governments,” Ban said. “Too few of them are willing to show courage and stand up for the principles of peace, solidarity and cooperation that underpinned the creation of the United Nations after the Second World War.”

Still, Ban said he remained an optimist, adding he hopes for the possibility of “transforming our political paradigms to support planetary peace.”

Choue In-won, chairman of the university system, stressed in his speech that change begins with practical civic consciousness, saying that global challenges can be addressed only when individuals recognize their shared responsibility and act collectively for a sustainable future.

AI raises new questions for education, security

At a press conference held ahead of the award ceremony, experts voiced particular concern over AI’s rapid development and spread, warning that the technology is advancing faster than governments, universities and civil society can develop effective safeguards.

Avi Loeb, a Harvard University theoretical physicist and a member of the Miwon Peace Prize selection committee, said AI tools are already approaching — and in some cases exceeding — the capability of highly accomplished human experts.

He cited the example of an atmospheric scientist who turned to Claude AI agents to assess criticism of his work, describing it as a sign that research and higher education must reassess how they operate in an era of increasingly capable AI.

“Science is now at a different place, where AI systems are as capable as the most accomplished experts,” Loeb said. “That is a whole new landscape that we have to operate in, in research, in universities. And it’s a huge challenge.”

Loeb cautioned that seemingly stable institutions cannot by themselves guarantee safety from catastrophic risks. Referring to AI developers who have publicly discussed a possible, though uncertain, risk of human extinction, he said even a small chance of such an outcome should prompt stronger oversight.

“If you were crossing the street and you had a 10 percent chance of being run over by a car, you would not cross the street,” he said. “Ten percent is much less than 100 percent, but it’s not negligible. So it looks to me like we always are on the brink of destroying everything we have.”

Daniel Holz, chair of the Bulletin’s Science and Security Board and a University of Chicago professor, said AI could make nuclear security more unpredictable by accelerating decision-making and generating false or unreliable information. He warned that if governments or military systems rely too heavily on AI during a crisis, errors or “hallucinations” could increase the risk of miscalculation in situations involving nuclear weapons and other high-stakes security threats.

“We don’t want civilization to be dependent on individual luck,” Holz said.

The experts said humans must remain “in the loop” as AI becomes increasingly autonomous and capable of self-improvement. Drawing a comparison to parenting, they said intelligent systems cannot be assumed to act in line with society’s values without sustained human oversight.

“You want to align them with your set of values. That’s the whole point about education. So how do you achieve that? Well, you achieve that by keeping yourself in the loop and not letting the system run wild,” Loeb said.