The ISP already knows the IP assigned to the PC that gets online through the ISP's service, and potentially every connection it makes, up to the point where it connects to a proxy.

So what would be gained by the "digital ID"? Well, maybe you hope to distinguish Alice from Bob when they both use the same IP. But how are you going to prevent them from using each other's credentials? The only possibilities are (a) motivating them by making their bank accounts or other values subject to compromise if they share their credentials, and (b) making the PC enforce the individual's approved logon with biometrics. With (a) you're only creating a new target for phishers and a new trap for technically naive users. With (b) you're doing the opposite of security, you're abolishing the possibility of security because the owner is no longer in control of the PC.

The other evil effect would be placing everyone at the mercy of their oppressive governments, as paperghost points out. And what's it all for? Fighting malware? In that case you're solving the wrong problem. The cause of the malware pandemic is simply the prevalence of an unsecurable OS. And the solution is moving to default-secure, open-source software.

Now I expect you don't want to see this answer, and this comment will disappear, because your company depends on the defects in the products of that noxious monopolist. But the truth is, computer security must be maintained on every endpoint device, and cannot be helped by mandates that compromise users' freedom.