Seems to be a code similarity tool being promoted (by the authors, not the frequent spin by outsiders seen on papers) as an IP violation risk tool, despite no expert in IP law being involved in developing it, and, judging from the abstract (the Github and paper links fail to resolve!) not even the slightest consideration of IP law in constructing the risk score.

WashU computer scientists developed a platform to evaluate the prevalence of intellectual property violations by code language models