Application Security
7/15/2014
12:30 PM
Sara Peters
Sara Peters
Quick Hits
Connect Directly
Twitter
RSS
E-Mail
50%
50%

Active Directory Flaw Lets Attackers Change Passwords

Aorato finds way to compromise Active Directory and change passwords without being noticed by SIEM.

Researchers have discovered a way for attackers to access and change the password of a user's Active Directory account, without being detected log-based security tools like SIEM.

The researchers, from AD security firm Aorato, say this is a severe flaw partly because of the ubiquity of Active Directory and partly because it allows attackers to do something that they may not be able to do even with physical access to a user's machine.

As lead researcher Tal Be'ery explains, if someone goes to get coffee, walks away from his or her desk, and forgets to lock his or her screen, a ne'er-do-well can sneak by, get physical access to the machine, and do basically anything that the logged-in user can do ... except change the password, if that person doesn't know the current one. That's why this new attack Aorato describes is significant, according to Be'ery.

Obviously, if the compromised account is one that's used often, someone will notice the password has been changed as soon as the individual tries and fails to log in. However, Be'ery points out that if this is done on a weekend, an attacker may evade notice for 48 hours. Plus, attackers could go after a dormant account to stay under the radar longer.

As Aorato explains in its report, the attacker first uses a publicly available free penetration testing tool (like WCE or Mimikatz) to steal the NTLM hash from user devices -- an authentication component that resides by default on devices that connect to enterprise resources.

NTLM is known to be a bit of a security hazard itself, and therefore, lots of organizations log NTLM activity. So, as Aorata describes it in the report:

2. The attacker forces the client to authenticate to Active Directory using a weaker encryption protocol. At this stage, the attacker uses the Active Directory flaw where the encryption protocol relies on the NTLM hash.

This activity is not logged in system and 3rd party logs -- even those that specifically log NTLM activity. As a result, no alerts, or forensic data, ever indicate that an attack takes place.

3. The attacker proves its so-called legitimate identity to Active Directory using the weaker authentication protocol.

According to Be'ery, when Aorato responsibly disclosed this, Microsoft said they do consider it not a vulnerability, but a part of Active Directory's design. Be'ery says, "We argue that it doesn't matter. A flaw is a flaw is a flaw."

 

Sara Peters is contributing editor to Dark Reading and editor-in-chief of Enterprise Efficiency. Prior that she was senior editor for the Computer Security Institute, writing and speaking about virtualization, identity management, cybersecurity law, and a myriad of other ... View Full Bio

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
Page 1 / 2   >   >>
kobrien82
50%
50%
kobrien82,
User Rank: Apprentice
7/23/2014 | 6:39:06 AM
Time to move on?
At some point, perhaps it makes sense to simply move on from Active Directory itself? Even as more companies are moving to the cloud and starting to build out new environments in some combination of AWS/Azure/Rackspace/whatever, there still seems to be a de facto approach of shoehorning in AD to attempt to manage authentication and authorization. 

One perspective (previously written about on DR) is that this flaw is indicative of how Active Directory is simply not designed to deal with the new infrastructure stack that many companies are moving to. The combination of poor design choices and thinly supported logging creates a threat surface that AD struggles to address, especially as authn and authz becomes more distributed. Active Directory is a legacy tool, and vulnerabilities like this are good reason to reconsider replacing it, especially as it becomes less essential and more of a holdover from a different generation of on-premise architecture.
Robert McDougal
50%
50%
Robert McDougal,
User Rank: Ninja
7/18/2014 | 10:00:15 AM
Re: Microsoft: "not a vulnerability, but a part of Active Directory's design"
I believe they view this as a non-issue because it requires physical access in order to pull it off.  If I have physical access I can do pretty much anything I want, this is just another drop in the bucket.  At least that is my way of looking at it.  It could also be that they are unable to fix the issue if doing so meant unravelling large portions of Active Directory foundations.
GonzSTL
50%
50%
GonzSTL,
User Rank: Ninja
7/17/2014 | 3:59:01 PM
Re: Microsoft: "not a vulnerability, but a part of Active Directory's design"
Physical access = all bets are off. That's my story, and I'm sticking to it.
Marilyn Cohodas
50%
50%
Marilyn Cohodas,
User Rank: Strategist
7/17/2014 | 1:05:17 PM
Re: Microsoft: "not a vulnerability, but a part of Active Directory's design"
Strongly agree, @securityaffairs!
securityaffairs
50%
50%
securityaffairs,
User Rank: Ninja
7/17/2014 | 12:49:59 PM
Re: Microsoft: "not a vulnerability, but a part of Active Directory's design"
Hi Marilyn, the security issues related to the NTLM are well known, but I disagree the statement 

"not a vulnerability, but a part of Active Directory's design" 

if an attacker could impersonate a user and changes its password it is a critical security issue!
Dr.T
50%
50%
Dr.T,
User Rank: Apprentice
7/16/2014 | 11:21:17 AM
Re: NTLM Authentication Bypass
Yes, Interesting. Client machines do not need to be booted form the devices, for servers we can just manage that as needed basis.
Dr.T
50%
50%
Dr.T,
User Rank: Apprentice
7/16/2014 | 11:17:17 AM
Re: Active Directory
Agree, if the domain is compromised, there is nothing else left to compromise. Most organizations do not encrypt the data at rest, it is free way to the data. We would think there is layered approach to security and each time you pass one layer it gets harder to pass the next layer, that is not the case, it gets easier one you pass the firewall.
Dr.T
50%
50%
Dr.T,
User Rank: Apprentice
7/16/2014 | 11:13:31 AM
Person in the chair
 

There would always be weaknesses if the person in the chair would not pay enough attention to the security. If she/he left the computer unlock he/she most likely posted the password to the monitor or under the keyboard too, nothing would help after that point forward. Security is a journey and everybody needs to be on board for it to succeed.
GonzSTL
50%
50%
GonzSTL,
User Rank: Ninja
7/15/2014 | 4:00:17 PM
Active Directory
Speaking of Active Directory, there's this Network World coverage about Black Hat 2014:

ABUSING MICROSOFT KERBEROS


Here's the abstract for this session by researchers Alva "Skip" Duckwall and Benjamin Delpy: "Microsoft Active Directory uses Kerberos to handle authentication requests by default. However, if the domain is compromised, how bad can it really be? With the loss of the right hash, Kerberos can be completely compromised for years after the attacker gained access. Yes, it really is that bad." They'll demonstrate such a compromise in real-world conditions.
Marilyn Cohodas
50%
50%
Marilyn Cohodas,
User Rank: Strategist
7/15/2014 | 3:24:30 PM
Microsoft: "not a vulnerability, but a part of Active Directory's design"
Hmmm...Does that mean that Microsoft absolves itself of any responsibiity for the mischief that this "design" could possibly lead to?
Page 1 / 2   >   >>
Register for Dark Reading Newsletters
White Papers
Flash Poll
Current Issue
Cartoon
DevOpsí Impact on Application Security
DevOpsí Impact on Application Security
Managing the interdependency between software and infrastructure is a thorny challenge. Often, itís a ďdevelopers are from Mars, systems engineers are from VenusĒ situation.
Video
Slideshows
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2014-0607
Published: 2014-07-24
Unrestricted file upload vulnerability in Attachmate Verastream Process Designer (VPD) before R6 SP1 Hotfix 1 allows remote attackers to execute arbitrary code by uploading and launching an executable file.

CVE-2014-1419
Published: 2014-07-24
Race condition in the power policy functions in policy-funcs in acpi-support before 0.142 allows local users to gain privileges via unspecified vectors.

CVE-2014-2360
Published: 2014-07-24
OleumTech WIO DH2 Wireless Gateway and Sensor Wireless I/O Modules allow remote attackers to execute arbitrary code via packets that report a high battery voltage.

CVE-2014-2361
Published: 2014-07-24
OleumTech WIO DH2 Wireless Gateway and Sensor Wireless I/O Modules, when BreeZ is used, do not require authentication for reading the site security key, which allows physically proximate attackers to spoof communication by obtaining this key after use of direct hardware access or manual-setup mode.

CVE-2014-2362
Published: 2014-07-24
OleumTech WIO DH2 Wireless Gateway and Sensor Wireless I/O Modules rely exclusively on a time value for entropy in key generation, which makes it easier for remote attackers to defeat cryptographic protection mechanisms by predicting the time of project creation.

Best of the Web
Dark Reading Radio
Archived Dark Reading Radio
Sara Peters hosts a conversation on Botnets and those who fight them.