Eh, I'm gonna posit something other that what's proposed: that these articles are intended to raise someone's publishing record.

I'd say there's 2 reasons one might pollute the datastream:

1. Obfuscate competitor's data queries

2. Use "approved" articles as a tool to trick regulators in a grander series of fraud.