Thanks to advancements in A.I., one of the big trends in security is automated hacking—in short, software that’s built to out-hack the human hackers.

The Pentagon’s research agency DARPA launched a Cyber Grand Challenge project in 2016, with a mission to design computer systems capable of beating hackers at their own game. DARPA wanted to show that smarter automated systems could respond and develop fixes far faster than a human—in just a few seconds.

A white hat hacker might spend months or years to spot and fix critical vulnerabilities, but the machine that won the Grand Challenge proved its might in just a fraction of the time. The winner became the first non-human entity to earn DEF CON’s black badge, which is the hacking community’s equivalent of an Oscar.

Very soon, malicious actors will create autonomous systems capable of automatically learning new environments, exposing vulnerabilities and flaws, and then exploiting them for gain—or any other objective, which could simply be generalized mayhem.