Wall Street was stunned Monday when Stanley Druckenmiller sharply criticized his friend and mentee Scott Bessent in a Wall Street Journal opinion piece, arguing that the Treasury Secretary’s intervention in the U.S. bond market was a mistake. Druckenmiller contended that government buybacks to lower yields were unnecessary, citing high inflation and federal debt as the real drivers of rising yields. The Journal’s opinion pages operate independently from its newsroom.

Investors quickly shifted from shock at Druckenmiller’s broadside to questioning whether he used artificial intelligence to write the piece. After social-media observers ran it through AI-detection programs, Druckenmiller confirmed in an interview that he did use AI to help compose the op-ed. “I’m kind of proud of using it,” the 73-year-old investor said. He described relying on AI tools as akin to using a speechwriter or calculator, noting that he and his family office use major platforms like Anthropic’s Claude, OpenAI’s ChatGPT, Google’s Gemini, and Perplexity.

Druckenmiller wrote the piece over the weekend while vacationing with his grandchildren, following a turbulent week in bond markets. Last Wednesday, the Treasury Department announced it would at least double buybacks for long-term bonds, causing yields to drop sharply before rebounding. The incident highlights growing debates over AI usage in public writing and whether authors should disclose such assistance.