The NY Times posted a great post with a conversation between a writer and an email scammer who was trying to impersonate a friend in need of money. It was a hilarious read and I had a fun idea to cut down on these types of crime.
Email scammers have a limited amount of time, they have to send out the scam emails that are crafted well enough to sucker someone in, then followup in a manner that reflects the person they are impersonating in order to steal money from friends or family. If their time is occupied with more and more people responding to what are obvious scams they are going to be too busy to actually scam people (hopefully… or the increase in responses creates super scammers with super instincts to tell who is a real target and who is messing with them).
Step 1: Build a bot similar to the bot Nigel Leck built on twitter that argues against global warming skeptics (also a hilarious article). This is 100% automated and is basically a very low level AI.
Step 2: Build a Google Labs feature for Gmail so that you can mark an email phishing or scammy and this bot replies based on what it sees in the first email and subsequent replies. So just like the NY Times exchange it can ask silly questions, verify banking details, and just waste the scammers time over and over.
Option 2: Mechanical turk the process! Create a google labs app similar to the above, but instead it acts as a middle man and sends the response to be responded to by a real person on mechanical turk for pennies, and then back to you to approve to send. The end result is scammers getting hundreds of responses back with no idea which one to pursue.
A silly idea but might really mess with some scammers heads.