An example of CAPTCHA

We here at Social Source Commons Blog recently got an email asking us to reconsider our use of reCaptcha for spam management on our blog. reCAPTCHA is a tool that uses the CAPTCHA technique to verify that the person requesting to interact with the site (post a comment, register as a user, etc.) is, in fact, a human and not a spam robot (or “spambot”) looking for places to automatically insert links to Viagra resellers and porn sites.

It works by showing the user an image of a word usually slightly distorted but still readable. The idea here is that bots can’t interpret what’s shown on a picture the same way that humans can so while a human can see that the swirly picture says “following”, the bot just sees an image and can’t pull out the word that is being displayed because it isn’t actual text but rather an image of text. The user then types in the word that she sees in the picture thereby verifying herself as a legitimate human commenter or user. The CAPTCHA system works very well and as of right now (spambots are constantly becoming more effective), it is the go-to security measure for websites. Website administrators can let CAPTCHA catch spambots before they event submit content.

The e-Sullivan, an braille internet device for deaf-blind people

However, CAPTCHA has a big problem. What happens when the user trying to submit a comment is blind and accessing the internet with speech or other accessibility software? The software cannot read the image any better than the spambot so that the user cannot complete the puzzle and thereby verify their human commentor status. The user’s comment is denied even though the comment is valid and from a human. Even if the software was able to speak the CAPTCHA text aloud, this would still exclude those who are deaf-blind who access the internet with braille tools. (see The Equipment for Deafblind People Page for more info).

The person who alerted us to our blog’s default disabled discrimination asked that we reconsider using CAPTCHA programs like reCAPTCHA and instead focus on plug-ins that check comments for spamminess rather than ineffectively screen for humans.

We thought this was a valid concern and a valid alternative so we’ve stopped using reCAPTCHA and are going to see how our spam management and security are affected. Instead of reCAPTCHA, we are using Akismet which looks at comments’ content and searches for specific spam-related terms and users.

If we are to provide a place for Social Source Commons community and nonprofit tech conversations and information, then we need to address accessibility for all of our users as best we can.

What are your thoughts about spam management, security and accessibility? Many sites like Facebook and Twitter rely on CAPTCHA programs for spam security causing social media to discriminate against the disabled. Is there a solution (yet)? Or are those with visual and hearing impairments casualties of the war on spam?

Further Reading:

• How Would You Design a Captcha for the Deaf-Blind?
• Let’s Stop Inaccessible CAPTCHAs
• Wikipedia – CAPTCHA: Accessibility