Warren Buffett Warns AI Risks Echo the Nuclear Age: ‘The Genie Is Out of the Bottle’

Warren Buffett said the dangers posed by fast-moving artificial intelligence resemble the existential risks introduced by nuclear weapons, warning that uncertainty among AI leaders heightens the threat. Speaking in a two-hour special on CNBC, the former CEO of Berkshire Hathaway said, “Even the people that are smartest about it say they don’t know where it’s going… the genie is out of the bottle.”

Buffett drew a parallel to comments by Albert Einstein during World War II on the atomic bomb—an invention that changed global power dynamics without changing how people think about conflict. He noted that nuclear proliferation expanded from a single nation to eight or nine today, arguing that diffusion—rather than control—has been the historical pattern.

While Buffett said nuclear risk can’t be solved with money alone, he emphasized he would deploy his fortune instantly if it meant permanently removing even a few countries from the nuclear game. He added that nuclear weapons were “absolutely necessary” for the U.S. during World War II, underscoring the moral complexity that also surrounds AI.

The billionaire’s concerns are longstanding. In Berkshire’s 2015 annual report, Buffett warned of catastrophic threats from cyber, biological, nuclear, or chemical attacks. More recently, at the company’s 2024 shareholder meeting, he said AI has “enormous potential for good and enormous potential for harm.”

Buffett’s latest remarks come as he steps back from public appearances. In a November 2025 letter, he said he would no longer speak at Berkshire’s annual meeting; CEO Greg Abel, who took over on Jan. 1, is expected to author the firm’s annual letter and field shareholder questions this spring.

As AI reshapes markets and geopolitics, Buffett’s message is clear: innovation without foresight can carry civilization-scale risks—and history suggests those risks rarely remain contained.