This is the trouble with making everything entirely cloud based. It relies on your cloud provider being willing and able to provide a service to you. Neither are guaranteed.
As noted above, there is no reason why, with a little thought, these devices cannot be designed to operate locally. Yes, it might require the user to obtain a piece of hardware or software to control the devices once their cloud component has been taken offline, but that's not an insurmountable problem.
One thing that also concerns me about this iot integration with cloud thing is who has access to the data? The devices themselves may well store that data, and may or may not have security vulnerabilities, but if it uploads to a server somewhere off your network, who has access to that server? Does the company have adequate controls on their processes? Do the technicians just look up your data for shits and giggles (I've read this does happen), or is their access controlled? It's bad enough that your smart plug might be logging when it is turned on or off, and that someone might be able to use that data to determine if you are home or not, but what about cloud based CCTV cameras? They are sending video of you, your loved ones and your environment to services you have little or no control over. Sure the GDPR offers some protection, but depending on the company that may well be after the fact. Essentially, you would be able to take action against them after someone has already looked around your house via the CCTV.
If I had a CCTV system, I'd go for one that stored all data locally. It may well have a cloud option, but that should be just that, an option.
It should be noted, it isn't just cheap knock off systems that potentially have these problems. I can't find the link now, but I remember that a year or so back, technicians at Ring were found to be looking at user's videos from their doorbells.
Biting the hand that feeds IT
