Lecture 04

IIT Kanpur July 201
20 followers
1 Video View·Jul 26, 2024
[Music]
okay so uh today what we are trying to
do in this module in this lecture is to
complete the uh module two which is uh
cyber killchain the lockhead Martin
cyber kill chain uh we have covered the
seven stages of locked Martin cyber kill
chain uh namely reconnaissance uh
weaponization delivery exploitation uh
installation command and control and
eventually actions on objectives so we
have to wrap up uh by uh actually
discussing a few uh important things
about uh defense right so so we saw that
the idea of locked Martin cyber kill
chain is that uh you can uh stop any of
the stages of the Cyber kill chain and
that way you break the chain that the
adversary was uh trying to create and uh
in your in your organizational uh
Network and uh to do that uh you know
you have to have defense in depth that
is you have to first try to stop the
reconnaissance uh uh creating very
robust uh uh rules and and policy with
respect to social media presence uh data
leakage Etc as well as uh all the
internet facing uh systems or Services
should be hardened uh you try to if you
if you fail to uh stop the
reconnaissance then you try to uh stop
the weaponization and uh so you have to
have proper uh malware uh security you
have to have the ability to check what
is happening in your uh uh devices
inside your organizational Network that
people are using and connecting to your
network or you have to have good uh
antivirus on uh people's devices so
that's way you know you can stop that
part of of the chain however in case you
even failed to do that you have to also
have to try to stop the delivery so
delivery you can have better email
security spam filtering and filtering of
fishing emails you can actually
uh make awareness high priority of your
organization so that people don't uh get
into uh the fishing traps or uh you can
do cyber drills or fishing drills to uh
make people uh more aware and on their
toes uh or you can actually also Harden
your internet facing uh services and and
the operating system uh these
applications or middle Weare that you
are using to make your um Pres
in the internet uh if you fail in the
delivery then you have to try to stop
the execution and to stop the execution
you have to figure out how you can
detect when some uh unwanted uh binary
is being executed and how you can stop
them uh from being executed so you need
endpoint security you have you need
agents on the endpoint so that they can
stop uh this from happening then uh if
you fail in execution then you try to do
uh that in installation when uh the
persistence is being created so to
create the persistence what people might
do is uh um you know register it on the
uh registry or uh they can put this
binary into the startup folder various
other ways to do uh the uh persistence
uh through installation so you have to
have again inpoint security to stop that
now what you uh if you are fail to do
that then you have to stop the C2C or
command and control traffic so you have
to observe the traffic do network
monitoring to see whether uh some C2C
traffic is actually going to uh certain
uh IP addresses URLs uh you may have to
use some a IML technique to detect which
URL is actually might be a C2C URL or uh
you know you have to have threat
intelligence regarding IP uh uh various
IPS that might be part of the C2C or uh
command and control infrastructure of
the adversary and if you can if you stop
if you uh do not succeed in stopping the
C2C traffic then chances are uh you know
the the malware or the payload has
already been working in your system it's
probably trying to do privilege
escalation it's probably trying to do uh
um you know internal reconnaissance it
probably is trying to do a movement
through the network Etc so to stop that
you have to have other techniques like
for example if the uh if they're trying
to do data exfiltration then you have to
figure out how to do uh uh data loss
prevention uh you have to have detection
of uh large amount of data getting out
of your system which is uh animalis
which doesn't seem normal uh or
privilege uh if the you have to have uh
what we call Pim the uh you know
monitoring the privileged uh uh accounts
and privilege uh use usage uh or you may
have to do uh uh various network
monitoring to see whether there is an
internal reconnaissance or you have to
also see whether uh somebody's uh you
know trying to do uh encryption of your
devices so as soon as the encryption
starts if you could actually detect that
through let's say use of trap files and
things like that you could actually stop
that so so so the main lesson here here
is that it's a defense in depth you have
to continue to organize your defense uh
with the hope that you will stop it at
one of the seven
stages so this whole uh idea uh is uh
based on the fact that uh the defender
is Vigilant and Defender is self aware
in the sense that it is monitoring the
network it is monitoring the activities
in the devices in the entire network all
the devices and have their logs have the
events have all kinds of information uh
new files being created new logins uh
happening all that information must be
coming in real time to a um what we call
Sim the the security incident and event
management server and from there it
should generate alerts based on rules or
AIML uh models and then the there are uh
there is the whole remedi remediation
team uh uh sitting in the security
Operation Center they look at the alerts
they immediately take some action some
actions may be trivial some actions may
require quite a bit of work from them so
to do this uh you have to also train
these people who are actually working uh
in your Security operation Center they
should actually be able to recognize
signs of various kinds of intrustion uh
they should be aware of the modest
support India of various uh Advanced uh
persistent thread groups how do how they
work so that as soon as they see alerts
Etc if you do not have any specific AIML
technique or signature based techniques
to detect that this is uh let's say ap28
or ap3 or any kind of advanced
persistent thread group uh cannot be
recognized by the AIML model you may
actually have good eyes of the very
well- trained s so engineers to actually
do that so uh so the knowing the
patterns uh and knowing the uh uh the
various ways that the threats are
carried out by various threat actors
would certainly help and that's what uh
you know the threa intelligence analyst
training uh is all about that uh uh that
you might have seen in various places
being advertised as one of the
certification programs by let's say EC
Council and others uh so basic idea is
that you have to get ahead of the threat
actor you have to be always Vigilant and
always uh in the defense defensive
posture to actually stop him at any of
the stages uh that you you can now the
issue is that how does the defender
learn right so the defender has to learn
all these things by analyzing the
incidences so that's why you you will
have cyber forensic experts who actually
do forensics after
the uh incidents happen so incidents
happen uh well uh in in the in the S so
parland SSC terminology uh first of all
uh there are events events happen events
uh May uh may actually rise to the level
of alerts uh because uh you know it
looks anomalous or a series of events
May look anomalous and the a signature
based uh pattern matching might uh tell
tell the S analyst that there is
something that needs to be looked at so
it will generate alerts uh and then uh
if the uh if this for certain kinds of
alerts uh based on the threat
intelligence that you have uh put in
into your uh system uh you might
actually declare something as an
incident and if the incident Rises to a
level where uh things are you know uh
requires a lot of resources lot of
people getting involved stakeholders
getting involved then we say that's a
cyber crisis so so this is this kind of
uh things whether some when something is
declared as an incident when something
is declared as an event when and
something is a crisis uh is based on the
uh organization cyber security policy so
they have to uh uh think ahead that when
do I call something as event something
as alert something as incident something
as crisis now when any of these things
happen espe special incidents happen or
or crisis happens uh most probably you
have been compromised and if you have
been compromise there are various kinds
of indicators in various parts of your
system it could be in your network
packet race that you have uh stored if
you didn't detect it in real time then
you you have the packet stream stored
you have the um uh devices that have
that might have been compromised with
malware so you have to look at their
memory for in memory dump we have to
look at their uh disk uh file system and
uh figure out what H happened and and
through the network Trace you can
reconstruct how the malware came in did
it come through an emails or did it come
through as a as somebody visiting a
website or did it come through uh an USB
drive uh various things so so you have
to look at uh look at these things and
and then once you look at these things
you have to know how to block it right
so you might have failed this time but
you will learn from this that this is
how the uh how to block it so uh and
then also in case you have uh an
incident which eventually did not lead
to a you know large scale uh you know
compromise or or data exfiltration or
ransomware uh encryption and things like
that then we say that that incident has
been somehow uh you know an attack has
been darten but in that case also you
have to do the for and six to analyze
how it really happened which part of
your defense did not work properly and
why it didn't work properly was it the
uh the patterns were not uh there in the
in your system to recognize or was it
that your AIML model failed or is it
because your firewall rules are
inadequate or whether your email uh
filters uh which are supposed to stop uh
you know malicious attachments and so on
is not working so so this is the way
uh you have to get the uh your own
threat intelligence so as I said before
the threat intelligence can be collected
from various different sources so one is
your own learning from your own
incidences but also uh you might get
threat intelligence from various uh
vendors who actually uh uh collect
various incidences around the world and
convert them to threat intelligence you
might also get it from your Regulators
or your uh s the cyber uh Emergency
Response Team uh Collections and and
feeds so there are various ways to get
threaten intelligence and feed that into
your uh defensive equipments protection
equipments like firewalls and intrustion
detection systems and endpoint detection
systems and so on to continuously
upgrade this uh uh this software uh the
uh uh the network monitoring software
intution detection software your
firewalls and so on so forth in order to
in order for you to actually get this uh
uh thing ready for the next uh Onslaught
so that's the idea of uh Defenders now
the the thing that we uh talk about to
these days we do not say that we we can
never claim that you know my my
organizational system uh is 100% secure
nobody can claim that because uh there
would be certain ways for the
perpetrator to get into your system like
we discussed in stocks net case it was
was through the USB drives uh in case of
uh the attack on Ukraine it was through
uh uh through the email fishing email to
high level Executives so so you cannot
uh ever be 100% sure that you will be
you will be protected uh so having uh
the protection uh uh you know protective
devices or protective Hardware software
whatever you have applications uh
doesn't necessarily mean that you can be
you can res easy so nowadays we say that
we do not say it's a cyber secure system
we we try to make this system cyber
resilient so resilience is the term that
we discussed before resilience is the
term for uh you know when your system
can continue to function even in the
face of uh very you know strong uh
attacks in this case in other other kind
of things uh it may not be necessarily
attacks but resilience is for from you
know your ability to you know uh
function maybe with a lesser performance
or lesser lesser Effectiveness but at
least it should not be you know uh
completely crash or crash and burn or
something right so so it has to be uh
you know what we call a graceful
degradation of quality of service but
you should be at least uh you know uh uh
provide some functionality and in case
um you know if the compromise uh you
know uh does uh compromise uh the uh the
functionality then you have to very
quickly recover so as soon as the uh the
attack stops uh or attack is withdrawn
then you should be able to recover very
very fast and that timing that is
required to uh recover is kind of a
measure of your resilience right so if
you if you if your system gets attacked
and you are in operational for 3 Days 5
Days 7 days it's Bann that means you are
not very resilient if it takes a couple
of hours you are much better resilient
than the other one so that's the idea of
resilience and resilience is the only
way we can actually uh measure how
effective our cyber security is uh you
know we cannot have as I said zero risk
or zero attack or kind of
scenario and uh as I also mentioned
before that the AP the Advan persistent
threats groups are actually funded by
nation state and they're very
resourceful and they're very
knowledgeable so therefore they find
various types of ways to do the attack
and you can be uh rest assured that they
will uh sometimes uh you know penetrate
through your protections and therefore
you have to be able to know how to uh uh
function and how to actually recover
very very fast and some cases you may
have to respond so you have to be kind
of agile uh so that you can uh defend
faster so later on like in this course
we'll talk about uh you know uh the
measure of resilience or what we call
crr it's a cyber resilience uh you know
measurement that that uh you will you
will look look at and you will see that
there uh you know you can be various
levels of resilience you may be totally
non-resilient or you could be resilient
to some extent but the best resilience
comes when your system is adaptive your
system adapts to the onslaught of dos
attacks or system uh uh you know has
enough redundancy that it fails over to
another uh part of the system to
actually provide the functionality to
its clients or customers so adaptive
adaptiveness or adaptivity is the is the
name of the game in case of uh uh you
know you really want to be
resilient uh so so as I said that and
you must have read the read the paper
from locked Martin uh so this um idea of
killchain is that just one mitigation
breaks the chain uh so they think that
it is a Advantage for the defender that
uh the defender has to actually break
the chain at one U one stage uh whereas
the other guy has to the adversary has
to do all the stages but I do not see
that as as realistic because uh what is
happening is that uh you can not be sure
so you cannot have like okay I I will
Target this uh installation stage or
I'll Target this reconnaissance stage or
I'll Target the C2 stage C2 you know
command and control stage in order to
break the chain because uh you know you
do not have the uh exhaustive way of
doing them right so you cannot be sure
that if I just focus on uh let's say
command and control I will catch all
kinds of command and control traffic to
stop them right there are as I as I
mentioned mened before there algorithmic
uh URL generation there are there are
very uh you know very quick migration
from one IP to another uh for the uh for
the uh command and control server uh
there are various ways this uh this uh
uh adversaries can actually uh you know
uh fool your uh tools or uh abilities to
actually uh break the chain so uh so I
think that you have to do defense in
depth you have to arrange for defense at
every for for every stage so that way
really you do not get a whole uh you
know advantage and as it says here the
defender has seven opportunities to
break the chain which means it has to
have seven types of defenses right so it
cannot just rely on only one stage that
I will break that stage and then and
I'll get uh the mitigation so the
conclusion in conclusion uh we can say
that
Defenders can have the advantage uh if
uh you think it like you know in that
perspective that you have to just break
one stage and the others have to uh the
adversary has to do all the stages uh
but bigger issue is that you know the
defender has to uh do better
communication and uh mitigate risks uh
and when I say better communicate we
mean that you know you have to
communicate you have to always assess
the risk and later in this course we'll
talk about risk assessment and risk
mitigation uh also you have to also make
all your uh employees who are actually
using uh or in charge of various assets
cyber Assets in your organization they
have to know about this risks and they
have to know all kinds of wayte threats
can be unfolded on them so so that they
can they can uh do better make better
decisions uh while using these assets uh
you also have to um you know build uh
real resilient uh design so resilient is
now the actually name of the game in
cyber security
so um you make sure you understand uh
that uh you know making system cyber
resilient is more uh very important more
and more important these days uh and
then you have to measure uh and
continuously improve now this is often
not understood by many cyber security
teams that if they actually uh do a
onetime risk assessment then they
actually uh figure out what are the
possible threats uh they might might use
a kill chain or or other uh ways of uh
uh formulating uh how the threat actor
might uh work in their system and then
they actually uh put in the defenses
like uh they can put uh perimeter
defenses they can put uh uh endpoint uh
defenses they can put network monitoring
they can put AIML tools for intution
detection or malare detection uh they
might do all these things but and also
train the people and so on but they have
to continuously measure how effective
they are in their in their defenses and
then based on the measurement results
they should identify where they went
wrong in the in an attack or or where
they can improve in case they did not
face at least did not know about an
attack but attacks are happening all the
time now the question is how aware you
are or how Vigilant you are how how how
you're monitoring how effective your
monitoring systems are you will you may
or may not see the attack but in any
case if you want to actually do better
then you have to continuously improve
and to improve you have to measure you
have to measure how effective you are so
remember there is no such thing as
secure uh systems are only defendable
but not necessarily uh 100% secure so
you have to think differently than you
might have been uh doing uh because most
of the cyber security training that
people usually get uh there we talk
about network security or we talk about
uh application security or we talk about
operating system security uh we uh do
not necessarily talk about a very goal
oriented um you know highly resourceful
attacker and how they can actually uh
string together vulnerabilities that are
in your network maybe in your perimeter
defense maybe in your uh endpoints or
devices uh maybe in the people like by
doing social engineering on them so you
have to start thinking about that and
and in once you start thinking about
that uh you can get uh you know uh know
where you might have to improve in your
cyber security um uh you know posture
like is it is it the processes or is is
it the uh is it that you are not
investing enough uh is it uh that you
are not communicating uh the uh the
risks and and the threats to the team
and the leadership uh uh to get the
proper uh investment into cyber security
uh by the organization or Staffing um
you're not getting and then your
architecture whether your architecture
is strong enough and uh second thing is
that you have to know your threats so to
do something well uh especially in this
thing uh in cyber security you have to
know uh what are the threats and many
times we actually get um into trouble
because we did not think about the
threats the threats uh what can happen
how can somebody do social engineering
or how can somebody do uh um you know
uh unpatched exploit an unpatched
application or unpatched uh operating
system or unpatched uh service that is
running on the internet whether uh
somebody can exploit uh uh uh you know
unnecessarily open ports uh in you know
where some some processes listening that
you were not aware of or how people may
be unaware of uh uh you know logins that
are being attempted uh uh several times
uh uh and eventually succeeding and all
that stuff or whether your uh OTP Etc
are properly uh protected all this kind
of stuff you have to think about so the
last thing in this context is this
courses of action metric bricks as I
have been telling you that this is just
a model right so this is a one way of uh
organizing your thoughts and organizing
your uh uh you know uh plans to actually
do you know what how I how am am I going
to make my system resilient right so uh
what this uh Matrix is showing uh we'll
see similar Matrix Matrix in another
context when we go beyond the LM uh
cyber kill chain but we actually go into
like uh miter ttps or we go into the
unified kill chain and things like that
uh but in this case also this is
something to actually if you remember
this table that can be uh very good for
you because uh you see that uh uh at
various stages uh there are ways to uh
detect whether something is happening uh
there are ways to deny something from
happening uh you can uh there are ways
to disrupt that thing from happening
there ways to degrade uh the ability of
the adversary or their ways to deceive
the adversary into thinking that it is
achieving its goal but while you are
actually observing them and then you can
also destroy adversary but you will see
that in the MIT defend uh case this
thing will be slightly differently
formulated but they all uh you know uh
are giving you a good uh uh way of good
uh terminology and and sort of like
onology to actually think about how to
organize your defense and how to
organize your resilience right so for
example here we are saying that if you
want to know whether somebody is doing
reconnaissance now there are many ways
to do reconnaissance here they are
showing you actually um one possible
connaissance that is true by visiting
your websites uh organizational websites
identifying the uh the people their
their contact addresses and so on so
that they can do spear fishing
uh fishing emails Etc in a very targeted
manner uh because they also will put
those fishing email language and
everything based on that person so it's
very personalized so it's highly likely
that the person will trust that email
and and and click on some link and so on
so in order to know whether some uh
strange people are actually doing uh
visiting your websites and and trying to
uh scrape data Etc so then you can
actually uh look at that from the web
Analytics or you know APO log or engine
X log or whatever web server you have
you can try to see whether there are
some anomalous activities so that's for
detection if you want to
deny for you know reconnaissance you can
also do uh you know various ways like
Access Control list or firewall for so
that they cannot actually access certain
web pages so certain web pages may be
behind the authentication wall or some
kind of firewall Etc or for if they try
to automatically scrape your websites
then you can actually use web
application firewall Etc to do that uh
to stop that so that's a deny now there
is uh you could probably think about how
to disrupt this by maybe or deceive them
by putting in names of people who are
who do not uh belong to your
organization and and wrong uh addresses
or or something but usually that is not
done so that that's that has other
repercussions uh with respect to uh very
uh you know reputation of the company
and so on weaponization uh you know if
somebody is trying to weaponize um you
know your uh weaknesses uh so you can
actually uh use a network intrustion
detection system or nids to uh detect
that uh uh somebody some email uh some
email is coming with uh malware
attachments or you can also detect
whether somebody's sending a payload uh
through the network to uh Target one of
your web Serv and things like that uh
and if you are a little more advanced
you can also use Network intution
protection system and that will actually
provide you with not only detection but
may actually protect you by by uh
stripping off those uh payloads uh from
the or or or you know stopping those
packets and and send it for for6 uh you
know by by denying to actually deliver
those packet to the actual Target and so
on
now in case of delivery the detection
now delivery could be again in this case
as I said that this phases like
reconnaissance for example so to web
analytics we can only stop one kind of
reconnaissance right we cannot stop for
example people doing you know Network
scanning to find uh your um services
that are running you know with
vulnerability and so on so similarly in
delivery case also when they say that
detection requires Vigilant user they're
only talking about one type of delivery
that is probably through email fishing
or fishing or smishing Etc uh they can
you can also stop this delivery through
proxy filters so you can have uh uh you
know these uh web proxies or or you can
have the email proxies uh so that uh you
know you have some filtering before it
is delivered to the end user uh you can
also disrupt them by uh having inline AV
that is when the antivirus is not
waiting to detect whether something is a
malware after the uh payload reaches the
uh end point but eventually the payload
has to go through your network to
eventually go to the Target uh uh email
server or from from email server to the
endpoint email client of the user so you
can actually um observe the network
traffic and reconstruct the files that
are being delivered uh and and in case
uh you find that the file is uh uh is a
malware that is trying to be delivered
then you can stop it right so you do not
have to necessarily uh uh send this to
the user uh deliver to the user uh but
in case it's an encrypted traffic the
best way is to actually have web proxies
so that the termination point of the uh
encryption is at the proxy not not all
the way to the end user so so at the
proxy the uh traffic will be unencrypted
and then you can apply AV on the traffic
if the traffic contains malware
degradation you can do by actually
slowing down suspicious traffic right
not necessarily stopping them but you
can create queuing delays uh so that you
know immediately something doesn't
happen and during the queuing you may
actually uh do also some kind of a
filtering now if you focus on the
exploitation phase what will happen is
that in the exploitation phase uh the uh
payload is already in the users's device
right so you have uh it has been
delivered uh and it is on the user
device so it's a host intution detection
system or what we call endpoint
detection systems nowadays uh that's
where uh that's where the exploitation
can be stopped if the intrustion
detection is uh good it can be detected
at least and once you have detected it
then uh you can you can do various
things but if it is detection usually
the host intution detection or endpoint
detection usually puts an agent on the
device and the agent reports all the
activities that are happening at the end
point to to a management server so by
the time the management server realizes
that the uh endpoint is has been has
been infected and and some file have
malware file has been downloaded Etc
it's it's probably uh lit in terms of uh
uh in in terms of stopping it but uh the
way to stop uh uh here way to deny is to
actually have a fully patched system uh
like your operating system middleware uh
you know farmw
or applications have to be uh fully
patched that is in there should not be
any known vulnerability that the users
might be able to exploit in case they
are able to execute or detonate the the
payload that doesn't necessarily 100%
mean that you will be able to stop
everything because as I said that there
could be zero day Zero click
vulnerability that could be actually U
you that may not be able to be fixed
even with a with a pasted version and
then uh to disrupt execution so there is
the DS data execution prevention so this
is one of the way to actually stop
buffer overflow attacks right it doesn't
necessarily mean that 100% d means 100%
freedom from buffer overflow attack
there are other ways to do buffer
overflow attack especially uh you know
uh through uh return oriented
programming through uh uh you know using
Heap overflow Etc so data extion
prevention is one way of you know
stopping the ex ution of binaries from
the stack the uh the program stack so
but it's listed because this is one way
of course this is not the exhaustive way
or you know 100% guaranteed but it it
can happen it can help installation phas
like that's where you are trying to
where adversary is trying to create
persistence by making the binary it has
downloaded on your device uh to actually
write itself onto the startup folder or
or you know uh putting it into the
registry uh auto run registry key and
all that stuff so in that case again the
HIDs will help in detection if you say
that okay anything that comes other than
this set of uh binaries any other binary
that comes has to run inside a inside
some kind of a environment uh that is
very limited uh that does not have
access to the entire file system and so
on uh you can do that so that's what is
called a change root jail so you can
actually try to deny this binary or
payload to get access to all your files
and everything if you put it in a change
D jail but usually that is very you know
problematic for uh users uh because uh
many of the binaries they want to run
are not necessarily malware they have to
run it especially developers so in that
case you know putting everything in
change root jail may be a little bit uh
bad uh um you know problematic for the
for the user and disruption is uh when
the antivirus stops quarantine a binary
or payload and does not allow it to be
executed in that case you know you can
you are disrupting the installation
similarly in the C2 that is uh command
and control you can detect the command
control traffic by having a network
intution detection system uh you can
deny the uh C2 traffic uh uh through uh
configuring the firewall very quickly
adaptively or using uh also some access
control system uh uh also you can also
have um you know uh what do we call um
the micro segmentation or segmentation
of the network so um so you can
actually have certain parts of the
network in so isolated that it is very
difficult to establish a communication
to command and control from there uh but
yeah I mean this is uh the Deni denial
and then if you want to disrupt the C2
traffic then you can actually have an
Network intution prevention so the
prevention and once it detects that this
traffic is C2 traffic it it just doesn't
it's not like it just reports that but
it
actually effectively try to block that
traffic by uh you know uh configuring uh
rules or or firing certain functions so
that's nips tarpit is actually very
similar to what we saw in queing so it's
a degradation uh method that you will
actually slow down the traffic you know
so that you you do not have the kind of
traffic that the C2 server needs to have
kind of uh uh traffic uh the packet rate
Etc uh can be disrupted so the C2
effectiveness of the C2 server will be
uh at least in communicating with the
with your device uh the malware that is
now currently in your device will be
actually degraded you can also deceive
by doing a DNS degrade that is you can
actually have the DNS local DNS service
that is a local DNS resolver to
translate the URL into some known IP
address that you are controlling so that
the traffic really never gets to go to
the actual server controlled by the
adversary so that but usually the
malware nowadays are actually uh um
going to be aware of this trick and they
will try to uh do various tricks to
actually avoid uh you know uh they might
sell this when they see that you are
doing a DNS redirect and you may not be
able to actually do any kind of
forensics uh then actions and objectives
is is like you know you see that okay it
is trying to encrypt a device or it is
trying to uh move from one machine
through a weak protocol to another
machine and so on or it is trying to do
a privilege escalation so audit logs can
help by in detecting them uh it's very
difficult to deny because this is a
stage where everything uh is totally uh
the adversary adversary has gotten uh to
do everything in the all the six stages
and and it has succeeded so now the
payload is in your device so your device
it's very difficult to stop it uh you
know from uh doing things it may not
succeed if you have a hardened uh
application uh you know so privilege
escalation can be can uh can be stopped
by if your application has no such uh
vulnerability that allows privilege
escalation if your operating system is
totally hardened and and has no
vulnerability that kind of stuff can
actually stop uh uh the action on
objectives but this action on objectives
is kind of catch all in case of in the
lockin Martin you know uh cyber kill
chain so there are many things the
payload can do right so so depending on
what it would do then you know denial
process will be denial method will be
the related to that uh you can degrade
the quality of service so that you can
actually stop uh for example if if you
can slow down you can probably have the
malware not work as fast and thereby you
can actually um contain the damage you
get enough time to contain the damage
honey pots are the way to deceive so you
can actually do something like a
Honeypot orchestration by which you will
actually uh when when the attack is
happening you detect the attack but you
are not able to stop it so what you do
is you actually somehow redirect the
attack traffic to a Honeypot service
which is which looks very similar to the
service that is being attacked so the
attacker would then think that it is
actually working on on your device or on
your system and uh try to do other
things like bring in additional payloads
or communicate with other C2 servers and
so on and you you can get gain Trad
intelligence from that so that's what
hyot uh does so that's that's all about
this thing so other thing is that you
know if you have to see as I said that
you have to keep measuring for uh you
know Effectiveness right over time
whether you are uh you are doing better
right so so here is an you know it's
kind of a madeup example like
organization may want to see uh whether
in December uh was able to like detect
something weaponization but it didn't
stop there but it could stop the
delivery by having mitigation and so on
uh now in March it might find that it uh
was able to stop uh mitigate the
exploitation installation and C2 and
then in June it might say that it was
detecting ionization again and so on so
you can actually figure out that uh in
the December time you learned something
and uh you actually use that uh so that
is what they're calling leveraging the
new indicators so you get some
indicators of that uh like what things
were compromised what IP addresses were
in the C2 were used in C2 what malware
is used was used all this information is
called the indicators of compromise so
so you feed that into your protection
devices like your firewall Etc uh your
endpoint detection and so on so
hopefully next time if the same
adversary comes with the same
infrastructure
and the same set of uh you know Tech uh
Tech uh techniques and so on you should
be able to stop them but if it is a
different adversary different indicators
you may not be able to stop them right
so at least you may not be able to stop
them based on this learning in December
you might have other sources of threat
intelligence about other threat actors
that you might also have F fed into your
uh devices and from that you might be
saved from this other threat actor but
each thre actor has different set of
indicators of compromise so one threat
actor indicator of compromise may not
necessarily always help in stopping the
other threat actors but so this is a
this is an ongoing process for the
defense team of a of an organization
cyber defense team to actually continue
to learn new threat indicators new uh
indicators of compromise new uh ways of
uh attacking new threat models and and
then use that information very
effectively
if you do not use that information
effectively then it's not any good right
so that's the idea and and again this
locket uh Martin killchain cyber
killchain is actually a way of
organizing your thoughts and way of
organizing your defense uh in in
response to what the attackers might do
so it's a way of thinking or
conceptualizing what the attackers might
do doesn't mean that attacker is exactly
following these seven stages or even
like there are other things might be
outside the seven stages that the
attacker might do which maybe may not be
even be properly captured in this model
so it's a model think about it as a
model and way of thinking way of
thinking about the adversary and and the
threats and way of organizing your
defense uh you know based on the that uh
that uh framework uh we'll now uh move
on to the other Frameworks uh where you
will see more fine grain uh way of
looking at these things so that's where
we'll stop this lesson
[Music]
[Music]
[Music]
