Attacks/Breaches

3/14/2014
11:58 AM
100%
0%

Target Ignored Data Breach Alarms

Target's security team reviewed -- and ignored -- urgent warnings from threat-detection tool about unknown malware spotted on the network.

Target confirmed Friday that the hack attack against the retailer's point-of-sale (POS) systems that began in late November triggered alarms, which its information security team evaluated and chose to ignore.

"Like any large company, each week at Target there are a vast number of technical events that take place and are logged. Through our investigation, we learned that after these criminals entered our network, a small amount of their activity was logged and surfaced to our team," said Target spokeswoman Molly Snyder via email. "That activity was evaluated and acted upon."

Unfortunately, however, the security team appears to have made the wrong call. "Based on their interpretation and evaluation of that activity, the team determined that it did not warrant immediate follow up," she said. "With the benefit of hindsight, we are investigating whether, if different judgments had been made, the outcome may have been different."

[Collaboration with competitors may be the key to slowing security threats. See Retail Industry May Pool Intel To Stop Breaches.]

Target arguably wasn't breached because it failed to invest in proper information security defenses. In fact, Snyder said the company had "invested hundreds of millions of dollars in data security, had a robust system in place, and had recently been certified as PCI-compliant." Likewise, the retailer apparently heeded multiple warnings from US-CERT -- part of the Department of Homeland Security -- about the increasing threat of POS-malware attacks against retailers.

Unusually for a retailer, Target was even running its own security operations center in Minneapolis, according to a report published Thursday by Bloomberg Businessweek. Among its security defenses, following a months-long testing period and May 2013 implementation, was software from attack-detection firm FireEye, which caught the initial November 30 infection of Target's payment system by malware. All told, up to five "malware.binary" alarms reportedly sounded, each graded at the top of FireEye's criticality scale, and which were seen by Target's information security teams first in Bangalore, and then Minneapolis.

Image credit: Jay Reed on Flickr.
Image credit: Jay Reed on Flickr.

When reviewing Target's log files, digital forensic investigators also found the November 30 alerts, as well as multiple alerts from December 2, all of which tied to attackers installing multiple versions of their malware -- with the alerts including details for the external servers to which data was being sent -- Bloomberg Businessweek reported. Later on December 2, attackers began siphoning 40 million credit and debit card numbers from POS terminals, as well as personal information on 70 million customers. Ultimately, they exfiltrated at least 11 GB of data, according to Aviv Raff, CTO of Israel-based cybersecurity technology company Seculert, which found one of three FTP servers to which the data was sent. From there, the data was transferred to a server hosted by Russian-based hosting service vpsville.ru.

Obviously, had Target's security team reacted differently, they might have contained what turned into a massive data breach. But the security team didn't even have to be in the loop. The FireEye software could have been set

Next Page

Mathew Schwartz served as the InformationWeek information security reporter from 2010 until mid-2014. View Full Bio

Previous
1 of 2
Next
Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
Page 1 / 3   >   >>
Ritu_G
50%
50%
Ritu_G,
User Rank: Apprentice
7/11/2018 | 4:54:43 AM
Re: Deactivation of FireEye's Automatic Response
Their security system specialists obviously need to resit for their security courses and tests. Due to their poor choice of defense mechanisms on that fateful day, things had turned out like how the retailer wouldn't have anticipated them to. This is a costly mistake that the team could refer to as a learning point. Having invested so much for their security system setup, the retailer obviously had greater expectations of the team and they had most likely anticipated such an incident to hit them.
Ritu_G
50%
50%
Ritu_G,
User Rank: Apprentice
7/11/2018 | 4:54:11 AM
Re: Deactivation of FireEye's Automatic Response
Their security system specialists obviously need to resit for their security courses and tests. Due to their poor choice of defense mechanisms on that fateful day, things had turned out like how the retailer wouldn't have anticipated them to. This is a costly mistake that the team could refer to as a learning point. Having invested so much for their security system setup, the retailer obviously had greater expectations of the team and they had most likely anticipated such an incident to hit them.
rradina
50%
50%
rradina,
User Rank: Apprentice
3/24/2014 | 3:19:56 PM
Re: Deactivation of FireEye's Automatic Response
It certainly does.  My last employer has been using it since ~2004/5 -- before McAfee bought Solidcore.  Back then the employer was flagged for not having virus protection on their POS systems.  We had to constantly ask for a compensating control.  That left me with a poor impression of the PCI rules and those who conducted the audits.  It's similar when calling a support line that isn't staffed by trained and experienced resources.  They cannot truly understand problems.  They can only read a script and follow a yes/no logic tree.
Duke_Bauer
50%
50%
Duke_Bauer,
User Rank: Apprentice
3/24/2014 | 12:03:15 PM
Re: Deactivation of FireEye's Automatic Response
I believe this solution exists (McAfee Solidcore)
pfretty
50%
50%
pfretty,
User Rank: Apprentice
3/19/2014 | 4:04:28 PM
Happens far too often
Unfortunate, but the fact that they ignored the warning signs isn't a surprise. There is a dramatic need for a shift in culture. One would think the cost alone would be enough. On average attacks cost companies $11.6 million according to the 2013 HP Ponemon Cost of Cyber Crime report (http://www.hpenterprisesecurity.com/ponemon-study-2013).

Peter Fretty (j.mp/pfrettyhp)
rradina
50%
50%
rradina,
User Rank: Apprentice
3/18/2014 | 11:54:26 AM
Re: Deactivation of FireEye's Automatic Response
Locking them down assumes an OS security exploit was not used to install the malware.  I think it's been established Target's POS uses Windows.  I'll even go further and make an assumption that it's probably XP.

I'm not aware of any XP built-in solution to prevent a security hole being exploited to install malware.  If it's a remote attack vector, it'll typically involve a network service of some kind.  Most services generally have escalated privileges and if compromised, the hacker can almost always use them to gain root access.  

What Windows needs is a helper that monitors via read/write hooks and compares all file-system changes on system/software components with a dictionary made on the original system's image.  If anything is found out of spec, an alert is issued and the processes that use the corrupt image are terminated.  Further, such a helper also needs to scan DLLs and applications IN MEMORY to make sure they too are appropriate.  If not, the processes are terminated.  If an new process begins that's tied to an executable that's not part of the original image, it's terminated before it even finishes loading into memory.

Such products exist for XP and had they been using them, it would have been really tough to infect their POS systems even if a USB thumb drive was inserted.  Hackers would first have to figure out how to disable that software before exploiting the system.  Unfortunately this would require hacking the system so the protection mechanism can be hacked.  It's a chicken and egg scenario.  Certainly not foolproof but arguably difficult enough to perhaps convince them a company using such protection is not low hanging fruit.
SaneIT
50%
50%
SaneIT,
User Rank: Apprentice
3/18/2014 | 8:43:30 AM
Re: Deactivation of FireEye's Automatic Response
You would think that the POS terminals would be locked down as tightly as possible.  It's not like your cashiers should be installing anything on them but not knowing all the details it is possible that the application used the name of a Windows service or application.  
PaulS681
50%
50%
PaulS681,
User Rank: Apprentice
3/17/2014 | 7:16:26 PM
unacceptable
It just keeps getting worse for Target. To now know they had the systems in place that could have stopped this breach if they just used the system correctly is unacceptable. This just goes to show you that the best systems are rendered useless id people don't use them correctly.
rradina
50%
50%
rradina,
User Rank: Apprentice
3/17/2014 | 5:33:00 PM
Re: Deactivation of FireEye's Automatic Response
They should respond manually.  If the product constantly cries wolf, either the alert config needs review or the product needs to be replaced.  If that's not an option then they should push the alerts to Splunk and mine the noise for credible events that correlate with other intrusion events (assumes firewalls and other stuff are pushed to Splunk).  My point was automated responses might be tolerated for devices that aren't customer facing but you do not want call center devices, bank ATMs or POS systems downed by a false alarm that automatically removes a vital component.

As a side note, I still don't understand why a POS system could have ANYTHING new installed on it outside of planned events.  They shoud use white list protection or an OS that won't run unsigned apps (like IOS, Android or Windows RT).
hho927
50%
50%
hho927,
User Rank: Guru
3/17/2014 | 2:25:10 PM
Block botnets
Target IT dept fail many ways. 1) If Target blocked all connections to botnet centers, the malware could not send data out. 2) The HVAC vendor said they didn't monitor Target remotely, why did Target give them a corp/network account? 3) Target should not give that account full access to the POS. 4) Security,access auditing was ignored. 5) Ignored alarms. 6) POS should have a seperate network. Target tried to save money here.
Page 1 / 3   >   >>
Veterans Find New Roles in Enterprise Cybersecurity
Kelly Sheridan, Staff Editor, Dark Reading,  11/12/2018
Empathy: The Next Killer App for Cybersecurity?
Shay Colson, CISSP, Senior Manager, CyberClarity360,  11/13/2018
Understanding Evil Twin AP Attacks and How to Prevent Them
Ryan Orsi, Director of Product Management for Wi-Fi at WatchGuard Technologies,  11/14/2018
Register for Dark Reading Newsletters
White Papers
Video
Cartoon Contest
Current Issue
Flash Poll
Online Malware and Threats: A Profile of Today's Security Posture
Online Malware and Threats: A Profile of Today's Security Posture
This report offers insight on how security professionals plan to invest in cybersecurity, and how they are prioritizing their resources. Find out what your peers have planned today!
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2018-9209
PUBLISHED: 2018-11-19
Unauthenticated arbitrary file upload vulnerability in FineUploader php-traditional-server <= v1.2.2
CVE-2018-9207
PUBLISHED: 2018-11-19
Arbitrary file upload in jQuery Upload File <= 4.0.2
CVE-2018-15759
PUBLISHED: 2018-11-19
Pivotal Cloud Foundry On Demand Services SDK, versions prior to 0.24 contain an insecure method of verifying credentials. A remote unauthenticated malicious user may make many requests to the service broker with different credentials, allowing them to infer valid credentials and gain access to perfo...
CVE-2018-15761
PUBLISHED: 2018-11-19
Cloud Foundry UAA release, versions prior to v64.0, and UAA, versions prior to 4.23.0, contains a validation error which allows for privilege escalation. A remote authenticated user may modify the url and content of a consent page to gain a token with arbitrary scopes that escalates their privileges...
CVE-2018-17190
PUBLISHED: 2018-11-19
In all versions of Apache Spark, its standalone resource manager accepts code to execute on a 'master' host, that then runs that code on 'worker' hosts. The master itself does not, by design, execute user code. A specially-crafted request to the master can, however, cause the master to execute code ...