Startup Robust Intelligence is experimenting with AI and how to mitigate the risks posed by malicious AI. Robust Intelligence is responding to threats of attacks on commercial AI algorithms, with the first one being reported in 2019 by the National Institute of Standards and Technology. As AI continues to be at the forefront of new technology developments, opportunities for exploiting weaknesses are also increasing. Robust Intelligence seeks to catch malicious input by using AI before it can wreak havoc on a system.
Robust Intelligence has created a program that utilizes AI to outwit the AI that reads checks, which is an early application of machine learning widely used. The program was developed by CEO Yaron Singer, and automatically tweaks the code written on the check. This creates confusion in the commercial check-scanning algorithm and would allow a scammer to empty a targeted bank account by modifying a legitimate check and tricking the scanning algorithm. The company sells two tools, one that probes AI algorithms for weaknesses and another that automatically intercepts malicious inputs like a firewall.
Read More: This Company Uses AI to Outwit Malicious AI