Re: [keyassure] The SSL Proxy Issue

[Yoav Nir <[e-mail]>]

On Feb 19, 2011, at 8:16 PM, Phillip Hallam-Baker wrote:

> OK and I assume that this means that this would certainly not be EV compatible (unless the signer is breaching the EV requirements)
> 
> There are two use cases that I can see would drive this:
> 
> 1) Enterprises that are required to track internal communications. 

What do you mean by "are required to"?  Lots of times this comes from either a "security officer" or an auditor who says, "there's this tool in the toolbox. Why aren't you using it?"
The legitimate use can be that the company has some application firewall that blocks HTTP based attacks. Without an SSL proxy, you can't block the attacks if they're over HTTPS.



> 
> 2) Governments looking to perform covert surveillance.
> 
> 
> I only recognize one of these use cases as legitimate. The other is really an anti-use case (attack).

I agree. An attack is an attack, even if the attacker works for the good guys. Even if the attacker is doing it for my own good.

> The only difference I can see in these use cases is the 'covert' part. I think it is OK to have a hole, provided the users know that there is a hole.


Well, it's the employer's computer and the employer's SSL proxy. The trust anchor store is controlled by the employer through the operating system's remote management tools. The employee need never know, so I'm not sure if this doesn't fall under 'covert' as well.