Keeping Secrets from a Mind-Reading Adversary

Imagine that someone captures your brain, and then proceeds to read and tamper with every neuron in your brain -- all while you are thinking about a secret that you want to protect. Is any security possible against such a strong adversary?

Fortunately, this scenario remains science fiction when applied to humans. But this situation is all too common when we replace humans with computers: software that dictates the behavior of computers is routinely captured and analyzed by adversarial entities. 

This is a problem that has vexed researchers for decades. In this talk, we'll discuss recent research that has for the first time opened up mathematical approaches to solving this problem.

Speaker: Professor Amit Sahai, UCLA