As if people weren't scamming before AI, with the billions of dollars seized from the Chinese scam dungeons in Cambodia [0] where they had kidnapped people and tied them to computers (including a few deaths, and another wave of arrests just recently)
Scammers gonna scam. At least if they can use AI for it then maybe they'll kidnap fewer people
Chinese models are going to keep getting stronger. Maybe the Free World™ should increase their people's access to and education about AI too.
Better to strike at the root of the LLM-scam business model wherever we can.
"If you scam us, we'll scam, you back" ?
I too am unconvinced.
Why would the scammer go thru your detector?
Phone? The network could spot sources exhibiting robocall patterns. AI could analyze live conversations and warn the recipient on certain patterns.
Instant messaging? There's spam and abuse screening built into most IMs already. AI could [locally/privately] analyze conversations.
Email? Ditto; analyze conversations, a bit deeper than basic junk mail filters already do.
Unless you ban mentioning other IM platform or communication channel, they will always try to get you on something less monitored