Why researchers are worried about hackable AI

On this week’s Priorities podcast, researchers at the National Institute for Standards and Technology warn of threats to artificial intelligence systems. Apostol Vassilev, a computer scientist at NIST, co-authored a report about how hackers can target AI through “data poisoning.” One of the biggest threats, he says, is that there’s no real defense against it. Also on the podcast, Joseph Thacker, principal AI engineer and security researcher at AppOmni, advises how state and local governments can protect sensitive information and still embrace AI amid concerns of data poisoning.

StateScoop’s Priorities podcast is available every Thursday. Listen more here.

If you want to hear more of the latest across the state and local government technology community, subscribe to the Priorities Podcast on Apple Podcasts, Google Podcasts, Soundcloud, and Spotify.