AI in research integrity and detection

Artificial intelligence is fundamentally rewiring the world we live in. There are good things, great things, and bad things about that. On the one hand, boring work can be automated, and creative work can be accelerated; on the other hand, mistakes can creep in unnoticed, and research misconduct becomes easier and more diverse. That creates new challenges in misconduct detection. There’s a serious question here: are we automating science, or automating research misconduct? One thing that’s no longer questioned is whether we need tools to deal with this problem. The irony is that the only tools with the adaptability required to detect all of these new things are, themselves, artificial intelligence tools. There is a new industry in research integrity screening, and these integrity screening tools already cover most of the world’s published literature.