They Ran a Secret AI Experiment on Millions—No One KnewWhat happens when AI pretends to be human—and changes your mind without telling you?In today’s episode, we break down the real Reddit experiment where AI bots secretly argued with 3.8 million users. No labels. No disclosure. Just data-driven manipulation designed to win. And it worked—6x better than real people.This isn’t science fiction. It’s a warning.We’ll cover:✅ How the bots were trained to persuade✅ Why the study violated basic ethics and trust✅ What the EU AI Act says about manipulation✅ What this means for marketers, leaders, and everyday users✅ Why this moment shaped the launch of my new companyThe age of AI is here. And the rules aren’t ready.Subscribe for more human-first AI insightsWhat’s your take? Drop a comment. This one’s personal.

Ask this episode

“What did the council decide on item 42?” — answered with the speaker and the moment.

Transcript

Related