The core issue with pig butchering is that the harm process is distributed across platforms, so there’s no single view of the problem at a platform level.

This hasn’t stopped efforts to improve knowledge and coordination, but the next barrier to action is in geographical location, and out right kidnap of people to man the fake accounts.

But now, to make matters worse, there’s LLMs which can simply fake humans at scale.

At that point, App makers would get access to the call and could do a variety of things with it.

So for instance, you might have a vulnerable grandparent you want to protect, so you install an aggressive AI screener app to handle these things.

Or, you might just send a log of calls to a loved one.

Or, you might send a transcript of that call to a loved one, or an AI, or both for protection.

Basically, Google and Apple saying "Potential Scam" says to me that they can know when a call is risky. They need to do more with that and open up an ecosystem of defensive software.

So before I call the number (say the phone number is 9), my phone app will log into a central server and say "My number is 7, and I'm trying to call 9". The phone 9 will check the server and see that "Oh I'm getting a call from 7" and then ask the server, "Is 7 trying to call me?", and will get a confirmation.

There'll be different warnings if 7 currently doesn't have data connection, or got a new phone and hasn't logged on for a while, or 7 doesn't even have an account on the server. If 7 has recently pinged the server and then tries to call 9 without notifying the server, the server can say "I can't see any sign from 7 that he's trying to call you." then 9's phone can warn her. There'll be different warnings for different situations.

And we can use certificates and do tricks like not needing an account for each phone number, but send a message "Yes this number from Microsoft tech support is actually trying to reach you.", signed with Microsoft's certificate. But then the scammers will get a certificate for "Micro5oft", etc.

Looking forward to a reply moaning about centralized authorities monitoring people's calls. Although as I write this I realize telcos can do this already. Perhaps it can be a P2P solution...