The Federal Trade Commission has one of the most potent tools to punish artificial intelligence-driven companies that disregard consumer privacy, and it’s not shy about using it. In recent years, the regulator has ramped up demands that companies destroy algorithms trained on “ill-gotten” data, signaling its growing urgency in trying to rein in AI.

“The FTC is tinkering with new tools,” Holland & Knight warns in a recent client advisory. “The FTC hopes companies violating consumer privacy to cheaply train their AI may think twice because the punishment is becoming more severe.”