I think they key is that data brokers are greedy and accept all inputs with little validation. Their data quality teams are more focused on algorithmicly removing “obvious” duplicates due to small transpositional errors and such. They are usually not geared towards stoping people from intentionally entering incorrect but valid data. Even in cases where they might hold back data because they are uncertain about, they don’t delete it, as soon as there are enough “signals” that it might be correct they pass it through.

So knowing that data brokers are greedy the trick isn’t to get governments to force them to delete and protect information, the trick is to give them so much bad information that the good data becomes indistinguishable.