Doomsday Clock expert calls AI an extremely pressing threat

By Splitfeed EditorialLast updated Sep 26, 2026, 19:17 UTC

On September 25, 2026, a member of the Bulletin of the Atomic Scientists, the organization responsible for maintaining the Doomsday Clock, issued a public warning describing artificial intelligence as an "extremely pressing threat" to humanity. The expert drew a direct comparison between AI development and the Manhattan Project, the World War II program that produced the first nuclear weapons and reshaped the global security landscape for generations. The warning was reported by USA Today, the Arizona Daily Sun, and Yahoo News, among others, on the same day.

The comparison to the Manhattan Project carries specific weight because the Bulletin of the Atomic Scientists was itself founded in 1945 by scientists who worked on that program and felt a moral obligation to warn the public about nuclear dangers. When someone operating within that tradition reaches for the Manhattan Project as an analogy, it signals that they see AI as a technology capable of producing irreversible, civilizational-scale consequences, not merely a disruptive economic or social force. What remains genuinely in dispute is whether that analogy is apt. Critics of such comparisons argue that AI lacks the singular, physically destructive mechanism of a nuclear weapon, while proponents of the warning would say the diffuse and accelerating nature of AI makes it harder to govern and therefore more dangerous, not less.

No published reactions from the Pro-AI, Anti-AI, or Middle Ground camps were available at the time of writing. Typically, Pro-AI voices would argue that comparisons to nuclear weapons are alarmist and ignore AI's capacity to solve problems of equal civilizational scale, from disease to climate change. Anti-AI voices would likely welcome the expert's framing and use it to press for immediate regulatory intervention modeled on arms control treaties. Middle Ground commentators would be expected to acknowledge the legitimacy of the concern while pushing back on the idea that the Manhattan Project analogy maps cleanly onto a technology with such a wide range of applications and developers.

The Doomsday Clock is formally updated by the Bulletin of the Atomic Scientists each January, so the next scheduled moment when this expert concern could translate into a symbolic shift in the clock's position is early 2027. Whether AI's weighting in that assessment changes, and how the Bulletin articulates its reasoning, would be the clearest signal of how seriously this warning is being treated within the organization itself.