Endpoint // Authentication
12/5/2014
11:35 AM
Keith Graham
Keith Graham
Commentary
100%
0%

Moving Beyond 2-Factor Authentication With Context

2FA isn't cheap or infallible -- in more ways than two.

What’s the best way to protect your proprietary business secrets or financial data? One common recommendation is to implement two-factor authentication. As you undoubtedly know, two-factor authentication limits the usefulness of any credentials that attackers may have acquired or created, restricting their ability to move laterally within the organization or access your VPN to log back in remotely and attempt to gain ever more powerful credentials.

Unfortunately, two-factor authentication isn’t cheap -- in more ways than one (or even two). Not only can it be costly to implement, it also disrupts legitimate user activity, increasing frustration and hurting productivity. Moreover, two-factor authentication isn’t infallible, as we now know thanks to the reports on the Operation Emmental attacks on Swiss and German banks, in which attackers scraped SMS one-time passwords off customers’ Android phones. 

Fortunately, there is another way keep attackers inside your network from getting what they want: context-based authentication. With context-based authentication, your organization can create rules that determine, pre-authentication, whether and how a given authentication process should proceed based on context.

What do I mean by context? Well, context can include information such as:

Device registration and fingerprinting
Device fingerprinting is typically a two-stage process: On first-time authentication, the solution registers an endpoint, and on subsequent authentications it validates the endpoint against the stored device fingerprint. The device fingerprint comprises a set of characteristics about that endpoint, such as:

  • Web browser configuration
  • Language
  • Installed fonts
  • Browser plug-ins
  • Device IP address
  • Screen resolution
  • Browser cookie settings
  • Time zone

Source IP reputation data
Your organization can deny authentication if the IP address of a user’s machine is part of the Tor anonymity network, a known botnet, or an IP or subnet associated with known bad actors.

Identity store lookup
In addition to stealing existing user credentials, attackers often create new ones. However, they often fail to create those users correctly, with appropriate group membership and attributes. Therefore, by comparing a user’s current information with the corresponding information kept in a directory or user store, you can thwart attackers attempting to use credentials they have created.

Geo-location
Context-based authentication can compare a user’s current geographical location against known good or bad locations and act accordingly. For example, users on a campus location can be approved while users attempting to authenticate from outside the campus can be denied.

Geo-fencing
Context-based authentication can also base decisions on a geographical area or a virtual barrier -- if the user’s location is outside of a certain proximity, then assign additional risk or deny the authentication attempt.

Geo-velocity
Using a user’s geo-location and login history together can also help prevent malicious access. For example, if a user logged in at 2 p.m. PST in California, it is reasonable to consider a logon attempt at 4 p.m. PST from the East Coast as an “improbable travel event” and deny the request.

Behavioral analysis
Over time, a solution can gather information about the way that a given user interacts with the device (such as keystroke dynamics, mouse movements, gesture and touch, and motion patterns). Later authentication attempts that fall outside the established behavior patterns can be denied.

Any of these techniques on its own could be circumvented, of course. But combining several or all of them offers a promising solution. Layering multiple contextual factors pre-authorization enables you to build a risk profile that can be used to determine whether to allow a user to proceed to actual authentication, deny the user's continuing with the authentication process, or “step up” to two-factor authentication. For example, if geo-fencing data, together with behavioral analysis, raises sufficient suspicion about a particular authentication request, rather than simply denying the request outright, the system can require the user to provide a second factor.

Context-based authentication can be tailored to your organization’s risk tolerance, enabling you to balance security with a better user experience. Users are unaware of the context-based authentication processes and are not burdened by two-factor authentication unless a login is deemed to involve a certain level of risk.

Attackers will get past even the most fortified perimeter, as newspapers remind us every day. But context-based authentication can help you keep them from them reaching the Holy Grail of your proprietary business secrets or financial data -- and keep your organization out of the headlines.

 

Keith Graham is Chief Technology Officer at SecureAuth Corporation. His expertise comes from 15 years in security, product management, product development, and consulting. As CTO, he leads product development and plays a major role in the creation and development of ... View Full Bio
Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
Page 1 / 2   >   >>
ODA155
50%
50%
ODA155,
User Rank: Ninja
12/9/2014 | 4:54:22 PM
Re: A Green Thumb
@Keith Graham,... Great article... but I really do hate to sound like a pessimist... but I believe you're talking about "locking down" the network, and when has that ever happened, that's a pretty big sale you'd be making, probably in the context of compliance, maybe?

"Fortunately, there is another way keep attackers inside your network from getting what they want: " Yes there is, instead of buying the "next big thing" or some silver bullet to secure our networks why not just hire the right people with the skills we need to properly manage all of the systems we already have? Seriously, think about this... how many systems do we have already running (fully implemented or watered down, mostly watered down) that were supposed to solve these problems for us and they probably could if there were enough people to do everything required?

I know it's pie-in-the-sky, but if (and I do say if... ) most companies large or otherwise had an admin dedicated to working with HR and management to manage user roles, privilege and access required to do a specific job... how many times does a user have local admin granted to install one requested\approved application and that privilege is never reclaimed... then we complain that too many users have local admin access.

I guess the reason this bugs me so much is because we're going through this very thing right now with a product that begins with an E and ends with an S and NOBODY has been happy for the last 11 months, also it was pushed off on the security department after IT started it and decided it was consuming to much of their resources.
Keith Graham
50%
50%
Keith Graham,
User Rank: Apprentice
12/9/2014 | 2:29:17 PM
Re: A Green Thumb
An example may be that in the case of source IP reputation data, an attacker could hop to a node that has a 'good' IP addresss, or an IP 'unknown' to any blacklists. Thats a really simple example. While you may think 'duh' of course attackers are going to do that, I would say that you wouldnt believe the audacity that alot of attackers have in terms of 'where' they 'come in from'.
Marilyn Cohodas
50%
50%
Marilyn Cohodas,
User Rank: Strategist
12/9/2014 | 1:35:52 PM
Re: A Green Thumb
Keith what are some of the gotchas in using 'context" in authentication? In what cases doesn't it work well?  
Keith Graham
50%
50%
Keith Graham,
User Rank: Apprentice
12/9/2014 | 11:39:35 AM
Re: A Green Thumb
Exactly right, and I always like to point out that while these methods on their own could be circumvented, the concept with context based auth is that these methods work together as a blend, with the idea that one or more may be able to trigger and slow down an attacker. Its about tightening that net as much as possible. I will also point out that at SecureAuth, we really do see customers using this as an effective detection and protection mechanism.
Tim.j.young25
50%
50%
Tim.j.young25,
User Rank: Apprentice
12/9/2014 | 11:32:09 AM
Re: A Green Thumb
OK I see, so an organization would just use some of the other factors you mentioned to begin (device fingerprint, geolocation, etc.) until a stronger behavioral pattern can be established. Pretty cool stuff!
Keith Graham
50%
50%
Keith Graham,
User Rank: Apprentice
12/9/2014 | 11:28:59 AM
Re: A Green Thumb
That's a good question Tim. All the methods of context that I've written about, with the exception of behavioral, will work out of the gate without the need for any learning period – in short, you're correct in the point you've raised. They're vendors out there today with behavioral capabilities who can verify a user, or flag on the risk profile of the user after a few days of user analysis. However, the reality here is that if an organization has an existing behavioral analytics solution deployed (thus a baseline already exists for a user), then that baseline can be used to form a risk profile and leveraged as just another method of context based auth, as part of the overall authentication solution.
Marilyn Cohodas
50%
50%
Marilyn Cohodas,
User Rank: Strategist
12/9/2014 | 9:16:57 AM
Locking down access
Does a janitor need access to the SOX regulated servers? Absolutely not, @Thomas Smith. Totally agree that identity management will be a major tactic in locking down enterprise security. But idea of adding various contexts to that identity is intriguing.
Tim.j.young25
50%
50%
Tim.j.young25,
User Rank: Apprentice
12/9/2014 | 6:13:25 AM
Re: A Green Thumb
Out of curiousity, how does an organization start off using context based authentication? For elements such as behavioral elements (keystokes, mouse movement, etc.) is there a sort of "learning period" to register how each user interacts with their device? 
Thomas Smith
100%
0%
Thomas Smith,
User Rank: Strategist
12/8/2014 | 11:27:38 AM
Re: A Green Thumb
Even if this article was to refer to biometric designs for 'fingerprinting', the article goes on to say you should be looking for more than any one method for doing two factors. I am a proponent of multifactor assessments for connections into an environment at ingress points. In my mind there are a ton of ways to secure a connection authorization that can and should occur at many different levels even prior to a user id and password being available to be input into any given connection state during the connection and authorization process. Using things like TPM indexed generated certificates, checked against your PKI issued infrastructure, to determine if a given machine even has authorization to connect to any given ingress point is a strong first method of authorization access. After which machine posture checks added with some security through obscurity and then user id/password + OTP authentication is a much safer and surer way to go. Using the concept of multifactor authorization process raises the level of comfort in knowing who is logging in from where and on what device.

Does this defeat everything? No, it does not. However the point of cyber security it not to defeat everything, we cannot defeat the human element. If a bad employee wants to hand over the keys to someone who will do something bad then no amount of security will prevent this. There is no end to the quest of a secure infrastructure; there is only higher confidence in the intelligence that what is being presented as true is true.

This is not to say it is a no win situation. We always need to look for the next level. What else can we do? Could we use our knowledge of the internet of things to determine how much access a user has rights to; not only based on the permissions a user has but also the place from which the user is logging in? For example if we reverse lookup the location of the authorization attempt and determine it's from the coffee house down the road then should that person have the same level of access as on the network internally? How about from a hotel located in a country known to have issues? Would you put it past some countries to put malware on the access points in a hotel to steal and listen in on the data flows?

For me the point of the article is to challenge us to think about what else we can do, what else can we define? Does geo-location role based security access make sense? Let me ask you, does a janitor need access to the SOX regulated servers? If not, then why do we allow this to, potentially, happen? We do not lock down access enough, not something as simple as security permission on a server infrastructure. What about the network infrastructure? I think we work very hard at doing a bunch of advanced things in our daily jobs, yet we forget the simple things all the time. Much like anything else in life if you do not constantly re-examine your own notions of what is acceptable then how would you grow?

 

 
Keith Graham
50%
50%
Keith Graham,
User Rank: Apprentice
12/6/2014 | 7:48:09 AM
Re: A Green Thumb
Joe - The article refers to device fingerprinting, not fingerprinting in the Biometric sense of the term. I agree that would be an entire other topic for debate!
Page 1 / 2   >   >>
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
5 Security Technologies to Watch in 2017
Emerging tools and services promise to make a difference this year. Are they on your company's list?
Flash Poll
New Best Practices for Secure App Development
New Best Practices for Secure App Development
The transition from DevOps to SecDevOps is combining with the move toward cloud computing to create new challenges - and new opportunities - for the information security team. Download this report, to learn about the new best practices for secure application development.
Slideshows
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2013-7445
Published: 2015-10-15
The Direct Rendering Manager (DRM) subsystem in the Linux kernel through 4.x mishandles requests for Graphics Execution Manager (GEM) objects, which allows context-dependent attackers to cause a denial of service (memory consumption) via an application that processes graphics data, as demonstrated b...

CVE-2015-4948
Published: 2015-10-15
netstat in IBM AIX 5.3, 6.1, and 7.1 and VIOS 2.2.x, when a fibre channel adapter is used, allows local users to gain privileges via unspecified vectors.

CVE-2015-5660
Published: 2015-10-15
Cross-site request forgery (CSRF) vulnerability in eXtplorer before 2.1.8 allows remote attackers to hijack the authentication of arbitrary users for requests that execute PHP code.

CVE-2015-6003
Published: 2015-10-15
Directory traversal vulnerability in QNAP QTS before 4.1.4 build 0910 and 4.2.x before 4.2.0 RC2 build 0910, when AFP is enabled, allows remote attackers to read or write to arbitrary files by leveraging access to an OS X (1) user or (2) guest account.

CVE-2015-6333
Published: 2015-10-15
Cisco Application Policy Infrastructure Controller (APIC) 1.1j allows local users to gain privileges via vectors involving addition of an SSH key, aka Bug ID CSCuw46076.

Dark Reading Radio
Archived Dark Reading Radio
In past years, security researchers have discovered ways to hack cars, medical devices, automated teller machines, and many other targets. Dark Reading Executive Editor Kelly Jackson Higgins hosts researcher Samy Kamkar and Levi Gundert, vice president of threat intelligence at Recorded Future, to discuss some of 2016's most unusual and creative hacks by white hats, and what these new vulnerabilities might mean for the coming year.