Re: Business opportunity
Encrypting universally is easy. It wouldn't even take VC money to do it.
Decrypting on the other hand...you might find some issues with that.
One of the many issues here is that key management is a fourth-order problem. You only need key management because you started encrypting (or signing, or authenticating) stuff. And you only started encrypting (or signing, or authenticating) stuff because other bits of the network of systems that handle your data don't otherwise adequately secure it.
And what's adequate? Well, that's up to you, and up to what your data is, and up to what the consequences are if your data is leaked, or copied, or corrupted. There's your first-order problem.
In IoT you have large numbers of actors, many different types of data, and crucially your expectations about who or what is allowed to decrypt any given lump of information may well change depending on very complex factors that cannot be analysed by the thingfrastructure itself.
Specific key management for specific closed, controlled scenarios is well understood, and has been for more than the decade you mention (though I concede even in these situatiuons it's rarely done right). Generalized key management to suit all possible definitions of 'adequate' for all possible combinations of dynamic actors, systems and data? That's the toughie.