Zero Trust Doesn't Trust You at All

At the tenth IDentiverse conference in Washington, DC, this week, enterprise security practitioners who deal with identity day in and day out met to find out the current status of the field.

While there are all sorts of specific solutions being hawked by their manufacturers, some companies have taken advantage of the gathering to spread what they think the future will look like from an identity-centric standpoint.

Rick Smith and Morteza Ansari of Cisco gave a presentation about Zero Trust (ZT) that started with a wider perspective than usually is found.

They think the security problem can be modeled by a popped kernel of popcorn, and that ZT tries to secure that popcorn. Once you are inside that kernel, you are all the way in. There really isn't a perimeter to it.

You can't make assumptions about authorization or access under ZT just by what the IP address says. Just because a request comes in on the enterprise VPN does not force you to automatically trust it, for example.

The Internet has a lot of signals that can help to make a dynamic decision about a particular access request.

In many ways, ZT inverts the current models by focusing of the trustworthiness of the user. The device being used in a request, the history of the user regarding the requested access can allow dynamic security decisions to be made. That dynamic range of choices available is central to the ZT ideals. ZT can allow differing base line analyses be done by enabling differing sets of transactions to be analyzed for different requests.

But Smith also made some very telling points. He said that, "We've done amazing well at identification and authentication over the last two decades. But in authorization we haven't gone as far. There is a lot more work that is needed… The signals used in making a decision on a particular transaction need to be the same for similar transactions for consistency. We basically have no way to deal with authorization when it crosses a security boundary."

ZT, despite vendor claims, is not a product in Smith and Ansara's view. It's an ecosystem with multiple products from multiple vendors all working together.

Paul Lanzi of Remidiant would agree with that. He works with Privileged Access Management systems that have a twist. He responded to what he saw attackers doing, which is go after privileged users. But the twist was adding a "just-in-time" factor. He co-founded a company in 2013 that focused on one problem with a solution that built upon past work (think NetWare) that had never been applied in this particular area before.

He told Security Now that, "Once we figured out that it was technically feasible to do that, we decided on it as a core concept of our company."

They keep Lockheed-Martin locked up, so they seem to have made some real word success in actually advancing security methods by being a part of a wider landscape rather than their own separate world. They work within the customer's ecosystem without connection to the deployed product.

While the conference is just beginning, the theme of vendors needing to cooperate and standardize is a big and pervasive one on the show floor.

— Larry Loeb has written for many of the last century's major "dead tree" computer magazines, having been, among other things, a consulting editor for BYTE magazine and senior editor for the launch of WebWeek.