I experimented with copy-move many years ago[1] after being inspired by a John Graham-Cumming blog post[2]. I got a job at Google specifically with the idea th

I experimented with copy-move many years ago[1] after being inspired by a John Graham-Cumming blog post[2]. I got a job at Google specifically with the idea that it would be the right place to experiment with internet-wide detection of falsified journalistic images, since it had the infrastructure and access to users that could support such an attempt. My actual job was YouTube ad infrastructure (gag), but I did have the chance to lead a team of 4 other engineers in a week-long hackathon where we created a version of Chrome that could automatically detect certain image alterations. It worked really well! We won an award, and the other engineers all wanted to continue working on the idea as their 20% time project. I imagined changing Google News or image search to offer up a small warning when images might have been modified to be intentionally misleading. Unfortunately I was terrible at increasing YouTube ad revenue, so my time at Google ended before I could try to implement the ideas on a larger scale. P.S. I've never seen Error Level Analysis give a useful result, ever. Maybe it can, but every time I've seen it used it gives completely ambiguous results that are spun however the person interpreting them wants. [1] http://lemonodor.com/archives/2008/02/protecting_journalistic_integrity_algorithmically.html http://lemonodor.com/archives/2008/02/protecting_journalisti... [2] https://web.archive.org/web/20160309195018/http://www.jgc.org/blog/2008/02/tonight-im-going-to-write-myself-aston.html https://web.archive.org/web/20160309195018/http://www.jgc.or...