What is the "Pesticide Paradox" in software testing?

A The observation that running the same set of tests repeatedly will eventually stop finding new bugs, as the software becomes "immune" to that particular test set B A rule stating that pesticides must never be mentioned in test documentation C A method for automatically generating new test cases from old ones with no human review D A principle that says testing should be done only once per release