Naive good intent or Psyops?
Google releasing "a large dataset of visual deepfakes" they've produced is - ostensibly - done with the aim of improving detection of deepfakes. However, it also provides producers of deepfakes (which, apparently, includes Google) a dataset against which they can refine their deepfakes so they no longer get classified as deepfakes.
Thinking about the "real" intent / motivation makes my head hurt: is Google optimistically naive, or is there some deeper psyop game behind it? Did a human make this decision or did some "ethical" AI come up with it?