Communications of the ACM
AI System Evolves to Evade Internet Censorship
University of Maryland computer scientists developed a tool called Geneva that automatically learns how to circumvent censorship.
Credit: University of Maryland
University of Maryland scientists led development of an artificial intelligence system based on genetic evolution that learned to automatically evade online censorship by repressive governments.
The Geneva (genetic evasion) system was tested in China, India, and Kazakhstan, and learned to exploit gaps in censors' logic and flaws that humans could not spot.
When operating on a computer that is sending out Web requests via a censor, Geneva tweaks the data's fragmentation and transmission so the censor fails to identify banned content, or cannot block the connection.
Geneva assembles sets of instructions from small code fragments, which follow refined evasion strategies for breaking up, configuring, or sending data packets.
Said Maryland’s Dave Levin, “With Geneva, we are, for the first time, at a major advantage in the censorship arms race.”
From University of Maryland College of Computer, Mathematical, & Natural Sciences
View Full Article
Abstracts Copyright © 2019 SmithBucklin, Washington, DC, USA
No entries found