Two types of dataset poisoning attacks that can corrupt AI system results

A team of computer science researchers with members from Google, ETH Zurich, NVIDIA and Robust Intelligence, is highlighting two kinds of dataset poisoning attacks that could be used by bad actors to corrupt AI system results. The group has written a paper outlining the kinds of attacks that they have identified and have posted it on the arXiv preprint server.

attacks bad bad actors computer computer science corrupt eth google intelligence nvidia poisoning researchers results robust intelligence science security server system team types zurich