Actually only AT&T is scheduled for a 2G sunset right now. Also I think one of the Northern European countries is choosing to maintain the GSM network and then retire UMTS. They're going to have a GSM for IOT and LTE for mobile communications. It's kind of insane. Wish I could find a source...
I don't think many GSM carriers allow internet originated, mobile terminated traffic, even for IoT. Mass compromise of IoT is going to be way less trivial for GSM capable devices than things on standard wifi.
What I'm trying to say is that while your quip is entertaining:
a) it's impractical to scan/compromise IoT devices remotely without a stupid design (by IoT standards stupid, btw- not regular internet design- we're talking REALLY REALLY STUPID).
b) I would bet my mortgage that the logic you are presenting was never considered by the relevant carriers. They're not that smart.
Not quite true. Verizon is planning to sunset 2G in 2019 [1], and is making efforts to get customers off 2G before that (similar to what AT&T has been doing). They are also rumored to be planning to shutter 3G in 2021. This leaves only T-mobile and Sprint in the US market providing longer-term 2G support, and their coverage is not great if you need to deploy devices outside major metro areas.
> They're going to have a GSM for IOT and LTE for mobile communications. It's kind of insane.
I don't think this is that insane. I doubt there are many devices that only support UMTS, but there are tons that only support GSM, including many in embedded systems that can't easily be upgraded. So by using most of your spectrum for LTE, you get spectrum efficiency, and supporting GSM you can support everything that doesn't do LTE (just not at high speeds).
