New Captchas Outwit Computer Spammers

May 25, 2009

(ChattahBox)—The war against the machines just received another weapon in its arsenal to fight against spam and cyber criminals. Website owners are waged in a constant battle with automated computer bot spammers that breach Internet security puzzles, called captchas, meant to distinguish humans from machines.

Once the spam bots crack the captchas, the machines immediately wreack havoc, signing up for email accounts to deliver spam and adding illegal postings that can be used to hijack Google search pages by blackhatters.

We all have experienced those distorted text puzzles when registering with a website, which often confound humans as well. Captcha is an acronym for “Completely Automated Public Turing test to tell Computers and Humans Apart.” And they pretty much do their job in protecting a website from automated spam programs.

Some websites have individualized their captchas, providing a real challenge to humans, as well as foiling machines, like the Quantum Random Bit Generator Service website, which requires people to solve a mathematical equation before allowing them to register with the site. This mathematically challenged writer won’t be registering with Quantum Random Bit’s site anytime soon.

However, it seems machines are becoming smarter and smarter and as soon as a new type of captcha is invented, a new software program is immediately written by hackers to crack the code and wage Internet warfare.

Bots can now crack captchas at a rate of 20 percent to create automated Hotmail accounts and at a rate of nearly 35 percent for Google Gmail accounts. Even at those low percentages, that amounts to a lot of spam and cyber crime.

Rich Gossweiler, a senior research scientist at Google, may have invented a new type of captcha based on inverted images instead of scrambled text, which just may foil the new crop of smarter machines.

Gossweiler and his team discovered that a human could easily figure out that an image needs adjusting to an upward direction, a task that’s impossible for a machine. The upside-down images are placed inside other images, making it even harder for machines to decipher. Human images are not used, because machines can easily recognize faces.

The new captchas can be fine tuned along the way, discarding images that humans have trouble with and ones that machines can identify. Users seem to like the upside down pictorial captchas, treating them as a fun puzzle and website owners benefit from a nearly limitless supply of images, which they can also tie into the website’s theme.

There is also a new service gaining popularity developed by the original creator of captchas, Dr. von Ahn, named reCaptcha, which farms out undecipherable pieces of text from digitized books and old radio shows, from public projects to archive human knowledge to make information more accessible to the world.

Websites using the reCaptchas not only secure their sites from spammers, but also contribute to a mass digitization of books written before the computer age.

The reCaptcha puzzles are sections of text or sound that the scanning device, Optical Character Recognition or OCR failed to decipher. Dr Ahn notes that people typing in reCaptchas decipher about 25 million words per day.

The audio reCaptchas are accessible to the sight impaired, but also are more secure than standard audio captchas that merely repeat the scrambled text. Dr. Ahn’s reCaptchas use audio clips from archived old radio shows, which are more difficult for machines to decipher, but more enjoyable for humans.

The reCaptchas are now used by thousands of sites, including Ticketmaster, Craigslist, Facebook, Twitter and The New York Times.

Someday in the future, machines may figure out how to recognize and invert images, but for now humans remain one step ahead of artificial intelligence programs.



5 Responses to “New Captchas Outwit Computer Spammers”

  1. IT Corner » Blog Archive » New Captchas Outwit Computer Spammers | ChattahBox News Blog on May 25th, 2009 5:52 pm

    […] the rest here:  New Captchas Outwit Computer Spammers | ChattahBox News Blog « Can you believe it? New Yorker cover done on iPhone app. LogLogic Raises $8.8 Million for […]

  2. Dennis Gamble on May 25th, 2009 9:08 pm

    Random guessing for an inverted image would have a 1-in-3 (33%) success rate, and 1-in-4 (25%) for a correctly-oriented image — not much if any improvement for all the extra effort to implement the new image CAPTCHAs!

  3. New Captchas Outwit Computer Spammers | ChattahBox News Blog « Tips on May 26th, 2009 12:42 am

    […] See m­­or­e h­er­e: N­ew­ Ca­p­t­cha­s Out­w­it­ Com­p­ut­er… […]

  4. Corporate Spam Levels Hit 90% In May - News: Everything-e on May 26th, 2009 5:27 pm

    […] a chief catalysts for the recent spike. Such a spike in May could also have to do with expected new CAPTCHA technology Google has now released. Bots became adept at breaking text-based CAPTCHAs, but the new form […]

  5. Corporate Spam Levels Hit 90% In May | Programming Blog on May 26th, 2009 7:57 pm

    […] a spike in May could also have to do with expected new CAPTCHA technology Google has now released. Bots became adept at breaking text-based CAPTCHAs, but the new form […]

Got something to say? **Please Note** - Comments may be edited for clarity or obscenity, and all comments are published at the discretion of - Comments are the opinions of the individuals leaving them, and not of or its partners. - Please do not spam or submit comments that use copyright materials, hearsay or are based on reports where the supposed fact or quote is not a matter of public knowledge are also not permitted.