Host: If you've tried to set up a myOrbit account recently, you've probably hit a bit of a wall right at the start. You're asked for a selfie for age verification before you can even see the first screen. And for a lot of people, that's a huge point of friction.

Listener: Yeah, I'll be honest, it feels a bit much. Most apps just let you check a box saying you're over eighteen and call it a day. Why is myOrbit being so intense about this?

Host: It’s because of how the product actually works. This isn't just a social feed; it’s a platform where AI agents are actually doing things—drafting emails, making phone calls, moving tasks around. Plus, there’s real money involved. Users can earn and spend, and payouts go to actual bank accounts.

Listener: Okay, so it's a high-stakes environment. But why not just do what everyone else does? Why isn't the 'polite fiction' of a checkbox enough?

Host: The problem is that on myOrbit, adults and minors are using the exact same app. It’s not a 'kids' version' and a 'pro' version. Because agents can act autonomously and adult content is just a toggle away in the settings, the company argues that a self-declared birthday is basically useless. It’s a legal paper trail for a courtroom, but it doesn't actually protect a fourteen-year-old.

Listener: So the selfie is the solution. But then we get into privacy. People aren't exactly thrilled about handing over their face to a new app they don't even trust yet. What happens to that photo?

Host: That’s the big promise they make: the photo is deleted immediately after the check. And they’re very specific about how they do it. They don't just 'hope' it gets deleted; it’s a guaranteed path. Even if the system crashes or the check fails, the deletion is triggered automatically.

Listener: Is there any world where they keep it? You know how these tech promises usually have a catch.

Host: They’re actually quite transparent about that. They won't say 'we never store the photo' because it has to exist for the duration of the check. If every safeguard fails, there’s a storage backstop that clears everything out within twenty-four hours. They basically say if a company tells you a file was *never* on their server, they're lying about how computers work.

Listener: Fair enough. What if the AI gets it wrong, though? I’ve seen those age-estimation tools mistake a bearded teenager for a forty-year-old or vice-versa.

Host: That’s where government ID comes in. They use Stripe Identity, but only as an appeal route. If the selfie check flags an adult as a minor, you can verify with an ID to fix it. But importantly, no content on the site is hidden behind a government ID requirement—it’s only used to correct mistakes in the age check.

Listener: I noticed you haven't mentioned *how* the check actually works. Is it analyzing wrinkles, or eye distance, or what?

Host: They actually refuse to publish that. They call it an 'adversarial system.' If they explained exactly what the AI is looking for, the people trying to bypass the check would have a roadmap for how to fool it. So the mechanism stays secret to keep the product safe.

Listener: It still feels like a lot of friction for a new user. It's a big ask for that first minute in the app.

Host: It is, and they admit that. But the trade-off they’ve chosen is that a moment of inconvenience for an adult is worth more than the safety of a child being compromised. They’d rather be 'early and imperfect' on safety than 'late and elegant.' If you want to see the full technical breakdown of their privacy backstops or how this fits into their broader safety policy, those links are in the report.