To live in the People’s Republic of China, or another repressive regime that purchases its technology, is to exist constantly under the eye of an all-seeing state. (For a panoramic look at just how thorough China’s internal surveillance is, check out Atlantic’s frightening piece on the Chinese “panopticon.”)
In China, the use of such keywords such as “Tiananmen” in an email or post on a website would surely trigger the automatic detection systems of state internet surveillance. There are many other words, topics and complicated ideas that trigger China’s state snooping and censorship. Now the University of Maryland, whose prowess in computer science and cybersecurity has been attracting admiration from around the world, has come up with a work-around for those who live in such repressive regimes and wish to discuss forbidden ideas or topics.
Two computer scientists at the University of Maryland have created an artificial intelligence software they call “Geneva.” This program is a heuristic learner, adjusting to the changing pathways of censoring technology and finding ways to close them off. Sometimes, Geneva instructs the censor to accept the forbidden keywords as benign. Other times, it tells the censor to move on – that the job is already done.
Most important of all, the user doesn’t have to install Geneva’s incriminating software. It can lurk in the background of a web browser until needed. Similar work is being done by the Open Technology Fund, a non-profit American corporation that has helped to create more secure communications for internet users around the world.
Given the deep interconnections of the internet, the capabilities of a repressive regime on the other side of the world should be a matter of great concern to us all. China can and does turn its eyes on American users. And given the growth of surveillance by U.S. law enforcement and intelligence agencies, we might someday need something like Geneva to protect us from threats to our privacy that are even closer to home.