We started the Bridging the Mobile Security Gap series by accepting that we can’t control the devices that show up on our networks any more. We followed up with a diatribe on the need for context to build and enforce policies which ensure that (only) the right users get to the right stuff at the right times.
To wrap up the series we need to dig deeper into enforcement, because as we all know the chain is only as strong as its weakest link. There are various places where mobile device security policies can be enforced – including on the devices themselves (via mobile device management) and on the network (firewall/VPN, IPS, network access control, etc.). There is no one right or wrong place to enforce policies. In fact the best answer is often “all of the above”. The more places you can enforce policy, the more likely your defenses will succeed at blocking attacks. Of course complexity is the obvious downside to multiple enforcement points.
Complexity has a strong negative correlation with operational consistency. You need to make sure your enforcement points work together. Why? Let’s run through a few scenarios where policies are not aligned. Yeah, they do not end well.
You can implement a policy forcing device to connect through the corporate VPN to receive the protection of the enterprise network – but that only works if the VPN recognizes the device and puts it in the right trust zone, with access to what the user needs. When that doesn’t happen correctly, the user is out of business – or a risk. Likewise, preventing misconfigured smartphones from accessing the network reflects good security enforcement, right? Sure, unless it belongs to the CEO who is trying to access a letter of understanding about an acquisition – even worse if you have no way to override the control. Exceptions are part of the game of managing security, so you need the ability to adapt as needed.
Both those scenarios result in users being unable to access what they need, which means a bad day for you. This is why neither MDM nor any kind of network-based control can operate in a vacuum. You can take a number of steps to attain operational consistency.
The first stop on our path to policy consistency is just making the enforcement points coexist. Do enough to make sure one tool is working contrary to the others. Unfortunately this is largely a manual process. Whenever changes are made or new policies implemented, your administrators need to run through the impact of these changes. All of them. Well, all the practical ones anyway. It’s a lot of work, but necessary, given how important mobile devices have become to business productivity.
Remember the good old days, when you did a similar dance when changing firewall rules. Some folks waited for the help desk to light up, and then they knew something was broken. We don’t recommend that approach. To avoid that problem vendors starting offering built-in policy checkers, and third-party firewall management tools emerged to perform these functions at higher scale and on multiple firewalls.
Unfortunately those tools don’t support mobile devices (or the relevant network controls) today, so for now you are on your own. That can be problematic, since you know (even if you don’t want to admit it) that it’s difficult to maintain operational discipline – particularly in the face of the number of changes made, exceptions managed, and other fires to fight. It’s not where you want to be, but coexistence is the start.
Integration at the console
The next step is console integration. In this scenario alerts funnel from one management console to the other. Integration at least gives administrators a coordinated view of what’s happening. It may even be possible to click on one console and have that link to a specific event or device in the other. Very fancy, and downright useful from an operational standpoint. A little less integration your admins need to perform in their own heads improves productivity. Of course this requires cooperation between vendors and these kinds of relationships are not commonplace. But they will be – enterprise customers will demand them.
Another benefit of this initial integration is more effective compliance reporting. Vendors map from a data source to the compliance report and pump the data in. That’s pretty helpful too – you know how painful getting ready for an audit remains, especially when you need to manage 5-10 different data sources to show to the auditor that you know what you’re doing.
Of course this is less than full integration – you still need to deal with multiple consoles to make policy changes, and the logic to ensure a policy in one tool doesn’t adversely impact another tool is missing. But it’s progress.
What you really want is the ability to manage a single policy, implemented across different devices and network controls. How cool would that be? But don’t hold your breath waiting. Like most other non-standards-based integration, we will see integration initially forced by huge customers. Some Fortune 50 company using a device-centric management product will want to implement network controls. They will call everyone together, write down on a whiteboard how much they spend with each company, and make it very clear that integration will happen, and soon. It’s the proverbial offer they can’t refuse, and they usually don’t.
Over time integration gives way to consolidation, and we expect MDM to be integrated into the larger IT device management stack and eventually work with network controls that way. Obviously that’s a few years down the road, but it’s the way these things work out. It’s not a matter of if but a matter of when. But without a crystal ball there isn’t much to do about that, so the best bet is to make decisions based on available integration today, and be ready to adapt for tomorrow.
Losing device specificity
We used to think of mobile devices as only laptops, but the pendulum has swung back the other way, to focus almost exclusively on smartphones as the greatest risk. But a device is a device is a device. It should irk you, as a security professional, that you still must use multiple consoles to do your job – what we call swivel chair management. Over time this integration becomes critical – it’s not like systems will get less complex, or you ever need to support fewer devices.
The long-term goal is to be able to enforce policies based on how your business works. Figure out who needs access, to what, from where, on what device, and when should they be allowed? That’s the vision of any computing, and it’s kind of a holy grail. We aren’t going to get there any time soon, but following the path to integration – at least for smartphones initially – will put you in a much better position to take advantage of the tools as they integrate over time.
And the tools will integrate. It just takes a lot longer than we like, or than guys like us expect.
With that we wrap this series. Please do let us know what you think in the comments; we will be packaging it all up as a paper over the next week or so.