Every day, people are already falling for AI fakes that look and sound 100% real. Tools can clone your voice with just 30 seconds of audio—ElevenLabs, Respeecher, anyone can do it. A 2024 University of Florida study found that people can only spot fake voices 73% of the time (meaning 1 in 4 times they think it’s really you). Keepnet Labs reported in September 2025 that 1 in 10 people have already received a voice-cloned scam call, and 77% of those who got tricked sent money.

Without a simple “This was made by AI” label, we’re forced to treat everything as possibly real or possibly fake. That kills trust fast and could have devastating consequences if it’s not regulated by the people now before it’s too late. One mandatory label fixes it: we still get the funny memes and cool art, but we shouldn’t have to question or look in the comments to debate if it’s real or AI generated . It’s not about banning AI because clearly it’s here to stay…it’s about not letting platforms trick us into confusing fiction for reality.