Trust Me, I’m Not Real (Yet): The AI Radio Host You Didn’t Know You Were Listening To

ARN’s hush-hush experiment with “Thy”—an AI DJ that quietly ran a four-hour weekday show on CADA for six months—has stirred up trust issues, ethical curves, and questions about what’s voice and what’s vibe these days.

Listeners began wondering, “Who is this Thy, anyway?”—no bio, no backstory, nada. Turns out she’s not human: ARN used ElevenLabs’ voice-cloning tech and based the voice on a real team member from finance. And yes, they did this without saying a word.

To get a sense of how people felt, ARN put together a survey asking listeners if they’d care whether a presenter was an AI, or if they’d feel betrayed being kept in the dark. The feedback? Pretty mixed, but rightfully uneasy.

So why does all this matter?

There’s a bigger picture here. We’re in this grey zone where authenticity feels like a luxury, and AI is creeping into places we least expect. A voice that grabs your attention—whether human or not—can still mislead. And that’s a slippery slope for trust in media.

Teresa Lim, vice president of the Australian Association of Voice Actors, didn’t mince words. She called ARN’s silence deceptive and emphasized the importance of transparency. Her push for AI content labelling laws? Not a bad shout, honestly.

Let’s wander off the beaten path…

You might think, “Maybe AI can streamline boring traffic updates or weather reads.” But when the personality starts to matter—when rhythm and rapport drive an audience—the absence of a real person becomes glaring.

Other broadcasters aren’t immune to AI temptations, by the way. A few in the U.S. and Poland have flirted with AI hosts, but many hit the rewind button after backlash. Still, ARN’s experiment is proof: it’s not science fiction anymore—it’s happening in real time.

A bit of real talk from me:

I get it—cost-cutting, automation, ‘push the boundaries.’ But somewhere, someone’s gotta ask: when did human creativity become optional? These AI voices sound real enough, but they lack the slip-ups, the warmth, the personality that tell us, “Yeah, this is lived experience.”

Maybe “Thy” was smooth and glitch-free, but the next time you hear an effortless radio voice, you’ll probably think twice. And maybe you’ll ask: is that person real, or did some coder hit ‘Generate Voice’?