Check Reference Links for Hidden AI Tracking Tags (Full Transcript)
[00:00:00] Speaker 1: So I had this paper from a researcher who was about to submit and, uh, she thought her references were looking good. And on the surface, they are looking good, but right here, here's the, one of the references that this pulled up that looks great, right? Not even visible to the eye, but, and even though the DUI checks out, you click this DII, it's a real reference, but lo and behold, scroll over somehow tucked in here, you have this little space here, UTM source, and I know this might be a little hard to see UTM source equals chat GPT, but if you've ever taken references from somewhere else and copied it in or gotten in your reference manager, things get changed where I moved around with track changes, this slips in and editors are doing checks on manuscripts now, because they're so worried about fabricated references, anything coming with these AI signatures, just not a good look.
- A reference can be legitimate while its attached link contains problematic metadata.
- The parameter “utm_source=chatgpt” may be hidden in a DOI URL and not visible at first glance.
- Editors may scrutinize AI-associated artifacts because of concerns about fabricated citations.
- Authors should review and clean URLs imported from external sources, copied text, or reference-management workflows before submitting manuscripts.