Introduction

Chatbots can be biased, deceptive or even dangerous. Today on “Post Reports,” we meet the hackers who are competing to figure out exactly how AI can go awry. 

Read more:

Will Oremus reports on technology for The Post, and recently that has meant writing a lot about AI and all the ways it could go wrong. 

“Even the people who make this stuff, the creators of these technologies, are also out there warning, hey, this could be really bad,” Will says. “This could go wrong in very disturbing ways.”

The range of potential harms is vast. And today, we meet the hackers trying to make chatbots go haywire. In what organizers billed as the first public “red teaming” event for artificial intelligence language models, we see a preview of Def Con, the annual hacker convention in Las Vegas – and we learn more about AI’s pitfalls.