turbocharged - 1:41 pm on Apr 28, 2013 (gmt 0)
We seem to have a problem with not just human scrapers, but of the automated variety. Watermarks in the case of automation would not work too well as I don't think these bots really care. It's a case of grab and post. It may impact the thieves bounce rate though when visitors see that the images are stolen, which could impact their ranks.
If Google does not read image metadata, especially the copyright portion, then they should. What Google is doing now definitely is not respecting the original creator when scraped content seems to rank so well.