A year ago, I warned about a little something I call The Random™️. It’s the wild animal inside every genAI bot that eventually breaks out and causes trouble.

The take-away lesson was that companies need to establish strong risk controls -- including keeping an eye on the bots and reviewing their output. Either that, or stick to using them in low-stakes environments.

Earlier this week, the LA Times offered up a concrete example of The Random™️ getting loose.

The 'Los Angeles Times' has removed its new AI-powered “insights” feature from a column after the tool tried to defend the Ku Klux Klan.

(Source: @TheDailyBeast on Twitter, 2025/03/04)

Longtime readers already know what I'm going to say here: laugh if you want, just don't forget to check your own work. Are you running a bot that needs more supervision?

For the full story on The Random™️, check out my Complex Machinery newsletter from February 2024, "It's still a wild animal."