A study conducted by researchers from Cornell University, UCLA, and UC Berkeley has revealed a concerning trend in scientific literature, estimating that around 150,000 fake citations were introduced by 2025. The research analyzed 111 million citations across 2.5 million research papers published from 2020 to 2025, focusing on platforms like arXiv and PubMed Central. The findings indicate that most of these fabricated references transitioned from preprints to peer-reviewed journals, with a significant rise in fake citations beginning around mid-2024, coinciding with the public release of AI tools like ChatGPT. Alarmingly, 78.8% of fake citations passed through arXiv's moderation, and 85.3% of those in bioRxiv preprints made it into final published versions. A separate audit published in The Lancet highlighted a similar increase in fabricated references in biomedical research, with rates climbing to one in 277 papers by early 2026. Researchers warn that the proliferation of fake citations could undermine clinical guidelines and systematic reviews, urging publishers to implement automated verification systems to ensure citation accuracy.