Transcript Its_time_to_go_hunting_-_IOC_vs_IOA_

It’s time to go hunting!
Indicators of Compromise
vs.
Indicators of Attack
Octavian Savin
Mihai Capraru
About Us
SecureWorks - global provider of intelligence-driven information security
solutions, which enable organizations to fortify their cyber defenses to
prevent security breaches, detect malicious activity in real time, prioritize
and respond rapidly to security breaches and predict emerging threats.
Mihai Capraru
• Security Systems Sr. Analyst at SecureWorks
• 5 years of experience in computer forensics and cybercrime investigations
Octavian Savin
• Security Systems Sr. Analyst at SecureWorks
• 4 years IT experience, 2 years security experience
Agenda
 The Cyber Kill Chain
 About Indicators of Compromise
 About Indicators of Attack
 Threat Hunting
 Case study: Unknown malware detection
• The Cyber Kill Chain
• Phases
1.
Reconnaissance
2.
Development
3.
Weaponization
4.
Delivery
5.
Exploitation
6.
Installation
7.
Command and Control
8.
Action on Objective
Indicators of compromise
 What are IOCs
 Sources for IOCs
 IOCs formats
What are IOCs?
Analyze
Data
 Pieces of forensic data found in log entries or
system files.
 Made off virus signatures, IP addresses, URLs or
domains, hash values, registry keys, filenames,
HTTP user agents.
 Created through a multi-step process driven by
analyst experience and knowledge.
Collect
Data
Tet
Create
IOCs
Identify
Affected
Systems
Deploy
IOCs
IOC sources
• Included in threat intelligence
feeds.
• Only released to paying
customers
• Rather expensive.
Commercial
and industry
sources
• Require a bit of detective
work and a good knowledge
of the internal network
• Potential to produce very
interesting results.
Internally
developed
IOCs
Free IOC
sources
• Specific IOC distribution
sites
• IOC Bucket
IOC formats: OpenIOC
• Open
source
framework
developed by Mandiant
• Utilizes XML to describe threat
information
• Easily transformed to a format
used by IT monitoring tools (Yara
and Snort)
• Free tools for managing them: IOC
Editor and IOC-EDT
IOC formats: CybOX, STIX and TAXII framework
Type
• CybOX
(Cyber
Observable
Expression) - provides a standard
for defining indicator details, known
as observables.
• Over 70 defined objects that can
be used to define dynamic events
or stateful properties.
• Objects: File, HTTP Session,
Mutex,
Network
Connection,
Network Flow, X.509 Certificate
etc.
Event
Action
Name
Associated
Object
Observable
Association
Type
File
Object
Process
Memory
Properties
Address
Library
System
GUI
IOC formats: CybOX, STIX and TAXII framework (cont.)
• STIX
(Structured
Threat
Information
Expression) - standardized, structured
language to represent cyber threat
information.
• Supports the following data elements:
 Observables
 Incidents
 Adversary tactics, techniques and procedures
(attack patterns, exploits, tools, infrastructure)
 Exploit targets
 Cyber attack campaigns
 Cyber threat actors
IOC formats: CybOX, STIX and TAXII framework (cont.)
• TAXII (Trusted Automated eXchange of Indicator Information) - set of specifications for exchanging cyber threat
information.
• Three sharing models:
Consumer
& Producer
Peer A
Subscriber
Subscriber
Producer
Peer B
Peer E
Consumer
Hub
Hub and Spoke
Consumer
& Producer
Subscriber
Source
Peer C
Subscriber
Peer D
Source/Subscriber
Peer-to-Peer
• In 2015, these standards have been transitioned to the Organization for the Advancement of Structured Information
Standards (OASIS), which has the role to develop them and promote their adoption, enabling cyber threat intelligence
to be analyzed and shared among partners and communities worldwide.
IOC formats: MAEC
MAEC (Malware Attribute Enumeration and Characterization) is a
standardized language developed by MITRE to describe information
about malware based upon attributes such as behaviors, artifacts, and
attack patterns.
• The MAEC Bundle – capture and share data obtained from the
analysis of a single malware instance.
• The MAEC Package – capture and share data for one or more
Malware Subjects (usually related). A Malware Subject contains
details of a particular malware instance and any other data derived
from analysis and metadata.
• The MAEC Container – enables a user to share any collection of
MAEC characterized data, including one or more Packages.
MAEC Container
(Tier 3)
MAEC Package
(Tier 2)
MAEC Bundle
(Tier 1)
MAEC Default
Vocabularies
IOC formats: MAEC (cont.)
• Cuckoo Sandbox can generate malware analysis
results in MAEC format.
• CybOX is supported in the latest version of MAEC.
• Benefits:
 standard representation – different tools can
handle the same malware data without data
conversion.
 improved data exchange flow - malware
analyses can be exchanged in a common, well
known and standardized language.
Indicators of attack
 What are IOAs
 Assemble IOAs
 Benefits of using IOAs
 IOC vs IOA
What are IOAs?



Series of actions that an adversary
must conduct in order to succeed.
All actions done by the attacker in
order to prepare his attacks.
All the “signs” left by the attacker in
earlier stages of the attack.
Indicators
of Attack
Assemble IOAs
Users activity
Vulnerability info
Suspicious
attempts
Network activity
Suspicious user
behavior
Servers/Host activity
Application activity
Security Devices
Analysis
Information
Gathering
(scans)
Database activity
Other …
Honeypot
Other…
Earlier stages
of the attack
Threat
Hunting
Alert
Benefits of using IOAs
Accuracy of detection
and response
• Provide indicators of action taken during every
stage of the attack.
Early
detection
• Help develop a strong game plan for a
company’s defense.
Faster
response time
IOAs
Damage
reduction
• Enable to understand the internal environment
and pinpoint probable targets for threat actors
Diminished
business
impact
Ability to see
attacks in
context
IOC vs IOA
• IoC’s are reactive indicators while IoA’s are
proactive indicators
• IoC’s can be used after a point in time, while
IoA’s are used in real time
• IoC’s are known, universal bad news, while
IoA’s only become bad based on what they
mean to you and the situation
IoCs
Malware
Signature
Exploits
IP Addresses
Vulnerabilities
IoAs
Code executions
User behavior
Malware behavior
Persistence
Stealth
Threat hunting
 What is threat hunting?
 Mindset needed in threat hunting
What is threat hunting?
“The process of proactively and iteratively searching
through networks to detect and isolate advanced
threats that evade existing security solutions, as
early as possible in the cyber kill chain”
Approaches
Frequency
driven by
Objective
driven by
Intelligence
driven by
Anomaly
Ad-hoc
Project based
Continuous
Mindset needed in threat hunting
• “If you know your enemies and know yourself, you will not be imperiled in a hundred
battles” (Sun Tzu)
• Three major questions we need to answer about attackers:
 What is their avenue of approach?
 What are they looking for?
 How are they going to get it?
• A formal hunting process has four stages:
 Hypothesis generation
 Investigation
 Pattern and TTP discovery
 Automated analytics
The Threat Hunting Reference Model Part 2: The Hunting Loop (28 Oct 2015). Retrieved from ‘http://blog.sqrrl.com/the-threat-hunting-reference-model-part-2-the-hunting-loop’
Case study
Unknown malware detection
Case Study – unknown malware
•
•
Process names:

cmd.exe;

wscript.exe;

regsvr32.exe;

powershell.exe
•
•
Host abnormal behavior
Windows legitimate processes used
with suspicious cmdline
Abnormal process tree
Connections created by unusual
processes
Suspicious command lines run
by the processes
•
•
•
Phase of
the attack
Benefits
IoA used
Parent name:

iexplore.exe
•
•
•
•
•
Endpoint Detection and Response
Full host visibility
Cross network searches
Initial stage of the attack
Attacker didn’t meet the scope
Earlier exploitation phase (not al
malware components were
installed/downloaded)
Tools
•
•
The files weren’t encrypted
The host can be sanitized not reimaged
The damage wasn’t produced
The incident was contained
Incident analysis - I
Process tree
Internet explorer child processes
- cmd.exe spawns
- wscript.exe spawns
- cmd.exe spawns
- regsvr32.exe
Incident analysis - II
Windows legitimate process used in stage I of infection
cmd.exe
1. Downloads malicious script from malicious site
2. Calls for wscript.exe to execute it
3. The script is downloaded as a .dat file, in the temp
folder
4. wscript.exe spawned
Incident analysis - III
Windows legitimate process used in stage I of infection (cont.)
wscript.exe
1. Executes the downloaded .dat file
2. Creates a .dll file
3. Deletes the previously used .dat file
4. Spawns cmd.exe
Incident analysis - IV
Windows legitimate process used in stage I of infection (cont.)
Second cmd.exe
1. Calls for regsvr32.exe to register the previously
downloaded .dll file
2. regsvr32.exe spawns, executes the command given
to it by cmd.exe
3. Followed by the creation of a ppp.key file
4. Finally communicating with the C2 server to receive
the encryption key
Root cause identification
1. The user searched something on
Google
2. Connected to the website
3. Redirected to the malicious website
4. Gets infected
Recommendations for ransomware mitigation
(based on previous findings)
Block process run in temp folder
Block identified malicious IPs and domains
Implement adware block solution
The end
Thank you!
Questions?