how will they detect hallucinated refs at scale? Manual spot checks? Automated DOI verification? The policy seems right but enforcement is the hard part.
how will they detect hallucinated refs at scale? Manual spot checks? Automated DOI verification? The policy seems right but enforcement is the hard part.
Enforcement is secondary and is allowed to take weeks / months / never at all if nobody reads the paper. It's about being able to ban if an issue arrises; not about keeping the database strictly clean.
However difficult it might be right now it's only going to get easier. Anyway I don't think proactive enforcement is the point. Rather now they have an official method by which to address incidents that are brought to their attention.