September 28, 2013 - A designer who previously worked with the NSA has created legible fonts that he claims cannot be detected by optical character recognition software, Kyle VanHemert of Wired reports. It's like those boxes where you prove you're not a robot, but used to communicate with other humans.
Sang Mun, who worked with America's signals intelligence agency during his time in the Korean military, said he was inspired by Google's attempt to mine almost all of the data that comes its way.
“Hearing about Google Glass — a 24/7 ubiquitous panopticon — the Google Goggles app and its new image search engine, and Google’s rigorous process of scanning every existing book ... All these software algorithms are programmed to extract every bit of information out of every kind of input,” Mun told Wired.
The typeface, dubbed ZXX , exploits weakness in recognition software to print messages that cannot be electronically parsed.
Mun also claims that a message using a combination of the four available fonts could be used to stifle a camera peeking over a shoulder.
After a year spent researching and creating, Mun released the type as a free download in 2012 .
Mun notes that the project is more of a provocation than a true security measure. Nevertheless, t he creation for a "defiant typeface" is justified.
This week a federal judge ruled that Google may have breached federal and California wiretapping laws for computer-scanning Gmail messages to create user profiles and provide targeted advertising.
The NSA also uses computers to search data for the identifying keywords or other “selectors,” storing the matches so that human analysts could examine them.
NSA legend-turned-whistleblower William Binney claims that the spy agency began using one of the programs he built, known as ThinThread, to map track electronic activities of Americans to collect "all the attributes that any individual has" and build a real-time profile based on that data.
Given the automatic data mining of human-generated data, fonts like ZXX may soon find a place in the digital world.
As VanHemert notes: "When you think about that sort of future, locating those vanishingly small edge cases where human ingenuity can foil machine intelligence becomes increasingly important."