Hooked on a feeling: The strange marriage of human rights and CAPTCHAs

Oct 23, 2012

About a month ago, the Swedish non-profit Civil Rights Defenders (CDR) released their version of CAPTCHA, the system that is used to verify that the user of the web site is a human (as opposed to a programmed robot). The verification is necessary because it keeps these “bots” from getting access to a particular section of a site and spamming it for information-retrieval and advertising purposes. Commonly, this involves asking a user to copy and submit a set of garbled or hard-to-read characters, something that a bot couldn’t do as easily as a human.

CRD’s alternative CAPTCHA solution, however, relies on the user’s ability to exhibit empathy. A user trying to pass the CAPTCHA must choose correctly from three different responses offered to the question “How does that make you feel?” preceded by a statement related to the issues in the organization’s global fight for the civil rights.

The Eastern Front

Out of 21 statements that should trigger an emotional response, 16 deal with issues in the Transitions coverage area. They include troubles that LGBT population faces given the ban of Serbian Pride parade, reckless homophobic public statements of Albanian and Montenegrin government officials, St. Petersburg’s homosexuality bill that bans “homosexual propaganda”Pussy Riot struggle in RussiaLukashenka’s Belarusian dictatorship and human rights offences in Kosovo and Chechnya.

Civil Rights Defenders’ head of communications explained in an email exchange that the focus on Eastern Europe “springs from the Helsinki committee on civil rights,” and that “we launched the CAPTCHA connected to Belgrade Pride, which is of interest to Eastern Europe”. She added that they believe the “CAPTCHA-format was a way to get [techies’] attention”.

For a more sophisticated CAPTCHA

Verifying that a web form is being submitted by a human being, rather than a bot, is a problem well researched in computing as spamming robots create unnecessary costs for maintenance. From solving a mathematical puzzle to deciphering a fragment of a scanned book, a CAPTCHA system requires human cognitive ability.

The problem is that these bots gets smarter all the time and human cognition and problem solving gets easier to fake. For example, some bots can solve CAPTCHAs through optical character recognition technologies. These technologies implement algorithms that can recognize text in images. They are extensively used in converting scanned books to digital formats where text is not an image but a set of characters.  That means that those CAPTCHA images that show text can be recognized and beaten by robots, rendering them obsolete.

Thus, it makes a lot of sense to explore the realm of emotions as this is something that machines still cannot replicate. Accordingly, the Civil Rights Defenders came up with a solution inspired by their own project work, but which ultimately comes with its own set of significant problems.

First, the CAPTCHA requires a dogmatic answer from a person in order to complete a mechanical task. This contradicts the very existence of technology, which is should be used neutrally as practical means to simplify and boost our problem-solving abilities. The technology, in the case of a CAPTCHA, is supposed to get out of your way by asking you to solve something that is not easily replicated by programming. Forcing the user to articulate a pre-defined emotion to gain access gets in the way of whatever else that user wants to do. And while a subjective opinion about politics rules out a programmed bot, it can also highlight potential differences among users themselves (assuming they have different opinions). This is definitely not desired when creating systems with a simple objective like a CAPTCHA.

Second, it forces a user to select one of three pre-determined answers, which fails to take into account the complexity and range of human emotion.

The culture of autism

The website boldly states that the system “takes a stand for civil rights issues”. It is questionable to what extent does one really stand up for human rights by trying to solve Civil Rights Defenders’ gullible CAPTCHA problem in order to sign up for a web service or complete a business transaction. Rather, it is more likely to denote another example of politically motivated slacktivism.

Given the open-platform nature of the Internet, the inevitable rise of false activism is mostly attributed to a change in what people consider learning and action. The modern Internet user’s activity is based mostly on gathering information, but spending the time to process the information or taking action is often skipped. Hence, solutions that merely inform a user—especially in this context, where the task complicates and interferes with the user’s activity—cannot be considered desirable from either a social or technological point of view.

Because programming to solve CAPTCHAs is still an extremely challenging task, humans still have the upper hand as their power manages to beat computing power in terms of operation costs. And since spamming is an industry in itself, it’s no surprise then that outsourcing exists in this industry too. In fact there are “virtual sweatshops” set up solely to use cheap, exploited labor to solve CAPTCHAs like these so that businesses can increase their spamming efficiency.

Unfortunately it isn’t hard to find the tragic irony of a virtual sweatshop worker trying to beat CDR’s CAPTCHA human-rights-aware questions day in and day out, all while working for a wage of up to a dollar for every thousand correct solutions.

About the Author

Ernad Halilovic

Ernad Halilovic is a TOL Editorial Intern. He studies Humanities and Computer Science in Prague.
  • Twitter feed loading...