A seventeen-year-old in Montgomery Township, New Jersey was charged this month after police say he generated A.I. nude images of his own classmates. The tip came through the National Center for Missing and Exploited Children. Three years ago, cases like this barely existed. Today, New Jersey investigators are juggling five or six at once. For a parent, that means the kid sitting next to your daughter in third period now has access to tools that didn't exist when you were in school.
According to a UNICEF, ECPAT, and INTERPOL study across eleven countries, roughly one-point-two million children had their images manipulated into sexualized deepfakes in a single year. In some places, that's one in twenty-five kids.
The law caught up. The investigation playbook didn't. Schools are still acting on digital evidence before anyone checks if it's real.
And when the fake isn't an image, but a voice in a boardroom, the stakes climb fast.
A finance worker in Hong Kong joined what looked like a routine video call with his colleagues. Every face was a deepfake. He wired twenty-five-point-six million dollars before anyone realized. That same week the Montgomery charges dropped, YouTube opened its deepfake detection tool to every actor, athlete, and musician in Hollywood. If you've ever taken a work call you couldn't quite verify, that uncertainty just became everyone's problem.
According to Reality Defender, the federal DEFIANCE Act, passed unanimously by the Senate in January, lets victims sue creators, distributors, and platforms for up to a hundred-and-fifty-thousand dollars per incident.
Detection isn't a feature anymore. It's infrastructure. And liability is shifting to whoever ignored the warning signs.
Which raises a quieter question, what about the voice on the other end of the phone?
Imagine your mother gets a call. It's your voice. You're in jail. You need money, fast. She wires eighteen thousand dollars before the panic clears. According to Trend Micro, one in three people who engage with an A.I. voice scam lose money, and Americans handed over more than five million dollars to these calls last year. Three seconds of audio from a birthday video is enough to clone you.
SQ Magazine recorded a four-hundred-forty-two percent surge in A.I. voice phishing attacks in 2025. The tools cost about sixty dollars a month. No technical skill required.
Experts now recommend a family code word. Not paranoia, the new locked front door.
Three stories. One pattern. The technology to fake a person, their face, their voice, their presence on a video call, is cheap, fast, and already inside our schools, our courtrooms, and our family group chats. Verification used to be optional. It isn't anymore.
Links to every story and today's podcast deep-dives are in the description. See you tomorrow.
