Teplitskiy: Can AI detectors tell whose ideas are in a research paper?

As scientists increasingly use artificial intelligence to write research papers, a new tool aims to determine whether a paper’s ideas come from a human or a machine. 

IdeaLens says it will analyze a science paper’s claims, arguments and evidence and help distinguish between research polished with AI and work completely generated by AI. 

University of Michigan School of Information associate professor Misha Teplitskiy, who studies how scientific knowledge is produced and evaluated, cautions that it may be difficult to understand what the detector is actually picking up. 

“It’s very hard to know what IdeaLens is really classifying off of,” he says.

RELATED 

Read “Did an AI or a human write that paper? New detector decides by examining its ideas” in Science. 

Learn more about Misha Teplitskiy’s research and teaching by visiting his UMSI faculty profile.