Imagine opening your phone and seeing a video of you being beaten, shot, or hanged, in your own clothes, in a familiar room, looking frighteningly real. It never happened, but your brain can’t shake it off. That’s the new face of online harassment powered by AI, and it’s crossing lines we never expected. This is exactly what happened to Australian activist Caitlin Roper. She saw AI-generated images of herself being burned or hanged, and what made it unbearable was a tiny detail: the woman in the photos wore the same blue floral dress Roper actually owns. “It’s these weird little details that make it feel more real… a different kind of violation.” From trolls to tech-enabled terror AI used to impersonate celebrities or make deepfake porn. Now, it’s being misused to create violent, personalised threats, making abuse more believable and far harder to ignore. Experts say this shift was predictable. Hany Farid, professor at UC Berkeley said: We will find exciting ways to use tech — and horrific ways to abuse it. What once needed special skills and hours of editing can now be done by anyone with a single photo and bad intentions. Why AI threats hit harder 1. They feel real: Words online can be ignored. But watching yourself “die” in a fake video hits the brain like a real threat. 2. Personalisation makes it psychological: Seeing your face, your voice, your home interior or your clothes makes the threat deeply personal, and traumatic. 3. It blurs fantasy vs danger: Victims can’t tell if the person plans to act or is just “creating content”. Cheap tech, maximum harm AI tools like text-to-video apps, voice-cloners and image generators have made violent deepfakes easy to produce. Some worrying incidents: Until recently, AI needed lots of pictures to clone a person. Now, one profile photo is enough. Voice cloning, once needing hours of audio, now needs less than a minute. Platforms under pressure but still failing Platforms claim they’re adding safety filters, yet users say guardrails are easy to bypass. When AI enters “swatting territory” AI isn’t just misused for images; voice cloning now powers fake emergency calls. Swatters generate audio of fake gunshots or cries for help that sound real enough to trigger police action. One school district went into lockdown over an AI-generated shooter call. What you can do for awareness self-protection Lock your digital footprint Reverse-search your identity Report fast — with evidence Know your rights Digital safety, a responsibility AI has brought creativity, innovation, and convenience, but also a new class of personalised cruelty. Threats no longer stay as text on a screen; they’re visual, realistic, and emotionally scarring.
As AI grows, so must awareness, safety tech, and digital laws. The world must learn to protect identity as carefully as it protects passwords.
The post Australian activist becomes latest target of AI-driven harassment:How artificial intelligence is making online death threats look disturbingly real appeared first on Tri-Cities India.
