Can AI be creative if it doesn't believe it can? We asked 556 humans and 13 LLMs to judge creative writing. Then we lied about who wrote what. Result: AI has learned to distrust itself, showing 2.5× stronger bias than humans against AI-labeled creativity. Proud of this work with @mmvty.bsky.social 👇
Wouter Haverals
📣 New preprint! We know humans are biased against AI-creativity. But what about LLMs, now often judging creativity in various contexts? Do they replicate, transform, or amplify this bias? We tested it. Turns out: AI is 2.5X more biased against its own work than humans. arxiv.org/pdf/2510.08831 🧵
