From my own experience, bots will actually help lock vandalization in through this process:

1. Article contains no obvious factual inaccuracies
2. Someone adds sneaky vandalization/lies deep in the article
3. Same person adds obvious vandalization, very soon after #2
4. Bot/human editor sees #3 and reverts it to #2, not to #1
5. Human editor adds something of value while not noticing the sneaky lies that are still there
6. Any attempts to revert vandalization at this point will be auto-reverted to #5.

Vandalization has now been locked in, and only very observing individuals that go through the edit history will notice.