Security vs. Accessibility

Today I read an interesting article from the W3C Accessibility Work group about the inaccessibility issues caused by CAPTCHA (Completely Automated Public Turing test to Tell Computers and Humans Apart)
Websites that provide access to multiple resources have to take precautions and make sure that the services they offer are actually used by individual human users and assure that their content is not harvested by web robots. In order to have this security issue covered CAPTCHA was developed by the Carnegie Mellon University (Pittsburgh, Pennsylvania). It usually consists of a set of distorted characters in a bitmapped image that the user has to recognize and type in order to avoid an inappropriate use of the web service resources. I always thought this idea was just great without thinking about further consequences. However, the WAI reminds us that things are not as resolutive as they appear. Obviously, these images don't have an alternative text value because this would give a way for robots to achieve their purpose. Therefore, this type of visual or textual verification means a great barrier for people that are blind, low vision or for those who have some kind of cognitive disability. In the article, somo alternative solutions are mentioned. Not all of them 100% valid, but it is definitely worth reading, if you hadn't done so before.

Finally, what they propose is the use of other methods for smaller websites that don't deny access to users with some kind of vision disability, such as heuristics checks or spam filtering depending on the purpose of the website. Furthermore they suggest to eliminate systems like the PINGUARD from ING Direct online financing, which may already be defeated by planned atacks and which means a short-term security benefit by sacrificing some people's autonomy to access their financial data.