After almost a year, I’ve decided to re-visit an old post I wrote regarding solving CAPTCHAs for cash. Specifically, people that want to use Google or Yahoo to spam, by automatically signing up for thousands of email accounts which requires humans to solve CAPTCHAs for them. According to MessageLabs, webmail based spam represents approximately 4.2% of all spam on the Internet - pretty significant.
There have been a number of articles on the Internet about automatic solutions to CAPTCHAs, but honestly, I find those stories somewhat dubious at best. Firstly, I don’t believe the solution rate is all that high as some people are claiming (it’s possible, but I don’t believe it’s happened for Gmail or Yahoo mail at the moment - if someone has actual proof I’d love to see it), secondly it’s super easy to change an algorithm to make it non-solvable again - keeping the automatic solutions at bay long enough to build another algorithm and so on. Lastly, there are very few people with the sophistication and know how to develop and use these tools as a percentage of the people who spam.
However, none of this issues deter a human CAPTCHA solver. If you remember my last article on this, we were seeing the economics drop significantly to where this is suddenly worthwhile, and if you read the comments of that post even more of these CAPTCHA breaking crews are popping up all over the world. Why wouldn’t they? Someone is willing to pay for it, so why wouldn’t you, if your family needed food? Sure the money may or may not belong to the spammer, but legit or not, the money is still real enough.
That leads me to something I found on the Internet while I was searching for more information on the economics of it. During my searching, I happened across some job offers for CAPTCHA breakers (also known as data entry). The advertisement was pretty intriguing:
The way the job offer is written is like it’s a stay at home sales person, or some other sort of semi-professional position. Words per minute, 12 hour shifts, a PayPal account along with an internet connection appear to be the only pre-requisites. I thought it was fascinating. Also, the economics appear to have dropped significantly from the last article I wrote a year ago. Now people are being paid $1/1000 CAPTCHAs solved, rather than five to nine times that, which is pushing this market into different directions due to increased competition. Perhaps there are other additional benefits for using a more expensive Romanian service verses the cheap version the Philippines are offering.
Unfortunately, I haven’t seen the operations personally, so I have to speculate that it’s less about the service and more about the cost of operations in the various countries. If anyone is willing to show me their operation I’d love to see it. In the mean time I think we should think about what exactly CAPTCHAs are offering us, and how we are sponsoring micro-economies in countries based on fraudulent human form filling. Is that really the goal? Is it actually the deterrent we intended? Perhaps we should be looking at other/better options.