Deepfake

  • |

    YouTube’s Deepfake Detector: Too Little, Too Late—or the Panic Boutique We Needed?

    ByHolden McGroin

    Picture this: the corkboard sneezed when YouTube quietly flipped the switch on its latest AI-powered tool mid-May 2026. The pandemonium machine—the one with a sales pitch notably absent of premium string—is half-alarmed, guarding faces but leaving voices wide open. You guessed it: the deepfake debate has entered your group chat.

    Here’s the newsflash: YouTube expanded its deepfake detection tool to users over 18, allowing them to scan for visual deepfakes potentially misusing their faces. As detailed by MWM, this feature employs a selfie-style scan via YouTube Studio, alerting users to any visual doppelgängers attempting to reenact their wild night as a ventriloquist. But there’s a catch—no shield for your voice yet, with promises of voice detection later this year.

    Just as the corkboard was settling in, ruffles of laughter echo as we learn this tool is opt-in. According to a Reddit report, users must enroll to be protected, raising the first eyebrow in our twitchy community of panic-chasers, where enrolling means facing the perilous task of finding the ‘Settings’ tab.

    Meanwhile, like a rumor with a ring light, audio deepfake scams are skyrocketing into the spotlight. As noted by TechRadar, one in four Americans received a deepfake voice call in the past year. Scammers are weaponizing AI, transforming a quick “Hello?” into an ominous “Who’s calling whom now?”

    While the visual detection tool offers a slice of solace, the true storm brews in our auditory channels. Yes, you can check if your face got cloned—but don’t answer the phone saying “Not my voice just yet.” We’re half-armored amidst an ongoing panic, a digital trench coat flapping in the algorithmic winds.

    So, even though YouTube’s new tool lets you shine a light on those visual pretenders, remember this: the real creeps might speak like you, not look like you. Let’s cling to the facts, fellow tinfoil enthusiasts, and perhaps keep a highlighter labeled ‘maybe calm down’ in hand.

    Sources

  • |

    FTC Cracks Down at Two Fronts: Uber’s ‘Cancel Anytime’ Scam vs. Deepfake Rescue

    ByLee Keybum

    The Federal Trade Commission has rolled in with a two-pronged attack that’s got consumers everywhere raising a hopeful eyebrow. First, they’ve locked horns with Uber over some dubious dealings with its Uber One subscription. Second, they’re clamping down on sketchy AI-powered deepfake abuses through the enforcement of the Take It Down Act. When tech platforms don’t play nice, the FTC’s bringing the heat—and perhaps your dignity and wallet back.

    In its latest one-two punch, the FTC kicked off with a May 5 lawsuit alleging Uber entangled users in its ‘cancel anytime’ Uber One promise, which was a bit like being told you could leave a locked room if only the door handle didn’t keep vanishing. Uber seemed to have misunderstood ‘unsubscribe’ as a feature only available when Mercury is in retrograde—or never. A transparent exit? That’s as rare as a well-behaved algorithm.

    Meanwhile, two weeks later on May 19, the FTC started flexing its muscles on the other front: defending against unwanted, intimate AI deepfakes with the shiny new Take It Down Act. Platforms now have less than 48 hours to take down non-consensual content. So, if the internet decides to wear your face like a cheap party mask, this Act is your public defender. Finally, a battle plan stronger than an AI’s wobbly moral compass.

    These moves are far from toothless. Platforms face civil penalties up to $53,088 per violation under these new rules, reminding them that failure to comply might further empty corporate coffers faster than you can say ‘user agreement.’ The FTC even preemptively fired off letters to major platforms to make sure no one’s caught napping at the duty wheel.

    On the upside for regular folks, there’s now hope that your subscription-induced déjà vu with Uber might finally end. And should someone decide to misuse your likeness, the FTC gives you a tool to demand action swift enough to make a cheetah look sluggish: TakeItDown.ftc.gov.

    So, next time you see the words ‘cancel anytime,’ remember—we might just be seeing that sweet escape become a reality. And as for AI’s attempts at playing Picasso with your profile, there’s a regulatory watchdog ready to prove there’s a better way to exist online than a digital free-for-all.

    Sources

End of content

End of content