There was this time when I noticed a colleague’s part of our manuscript were oddly inconsistent, as it didn’t match their usual writing style, and some of the references seemed made up. When we asked them about it, they admitted they’d used ChatGPT to write parts of it, including the citations, without checking if they were even legit or letting the group know. It ended up causing a bunch of delays since we had to go back and fix everything, making sure all the sources were accurate. It was a tough lesson about being honest and transparent, especially in research.