Science Is Drowning in AI Slop
2 months ago
- #Scientific Fraud
- #Peer Review Crisis
- #AI in Science
- Dan Quintana, a psychology professor, discovered a fake citation of his own non-existent paper in a journal he was reviewing.
- Phantom citations and AI-generated fraudulent papers are becoming widespread in scientific publishing, even in respected journals.
- Scientific journals, traditionally the gatekeepers of knowledge, are now struggling with an influx of AI-generated submissions.
- Peer review, established to manage manuscript volume, is now overwhelmed by the sheer quantity and sophistication of AI-assisted fraud.
- Paper mills produce fraudulent papers at scale, often recycling templates, especially in fields like cancer research and AI.
- AI can generate convincing but fake images and data, making fraud detection more challenging in biomedical research.
- AI research itself is now affected by fraud, with templates for fake papers and AI-generated peer reviews.
- Preprint servers like arXiv and bioRxiv are also experiencing a surge in AI-generated submissions, diluting scientific discourse.
- The rise of AI-generated content in science could lead to a scenario where most papers and reviews are AI-produced, embedding fraud deeply into knowledge systems.