WEBVTT

00:00:00.000 --> 00:00:03.270 align:middle line:90%


00:00:03.270 --> 00:00:03.880 align:middle line:90%
Hi.

00:00:03.880 --> 00:00:07.740 align:middle line:84%
Welcome to this video lesson
on Active Directory Tiering.

00:00:07.740 --> 00:00:10.140 align:middle line:84%
This one is the
first of two parts,

00:00:10.140 --> 00:00:13.770 align:middle line:84%
in which we will only focus on
the theory behind this model.

00:00:13.770 --> 00:00:16.890 align:middle line:84%
In the second part, we will
go and practically implement

00:00:16.890 --> 00:00:18.720 align:middle line:84%
the model in our
environment, where

00:00:18.720 --> 00:00:21.100 align:middle line:90%
we can showcase how it works.

00:00:21.100 --> 00:00:25.770 align:middle line:84%
Our short agenda includes, first
identifying what is the problem

00:00:25.770 --> 00:00:28.080 align:middle line:84%
and why do we need this
tiering model, what

00:00:28.080 --> 00:00:29.880 align:middle line:90%
does it solve for us.

00:00:29.880 --> 00:00:33.210 align:middle line:84%
And the real problem here
is the credential theft

00:00:33.210 --> 00:00:38.080 align:middle line:84%
that often occurs at endpoints,
and we will demonstrate that

00:00:38.080 --> 00:00:39.420 align:middle line:90%
just in a second.

00:00:39.420 --> 00:00:42.750 align:middle line:84%
And the second part of
the agenda for this video

00:00:42.750 --> 00:00:44.470 align:middle line:90%
is the model itself.

00:00:44.470 --> 00:00:46.590 align:middle line:84%
So we really drill
down into the model

00:00:46.590 --> 00:00:48.060 align:middle line:84%
and go through
different parts that

00:00:48.060 --> 00:00:51.890 align:middle line:84%
are really important to
implement it correctly.

00:00:51.890 --> 00:00:55.510 align:middle line:84%
So the problem itself
is credential theft.

00:00:55.510 --> 00:00:58.890 align:middle line:84%
What happens is that, when you
connect remotely or physically

00:00:58.890 --> 00:01:01.810 align:middle line:84%
at a Windows machine,
your credentials

00:01:01.810 --> 00:01:04.750 align:middle line:84%
are going to be stored
in the LSASS process.

00:01:04.750 --> 00:01:07.990 align:middle line:84%
And tools such as
Mimikatz, have the ability

00:01:07.990 --> 00:01:12.130 align:middle line:84%
to debug that LSASS process
and extract credentials out

00:01:12.130 --> 00:01:13.090 align:middle line:90%
of that.

00:01:13.090 --> 00:01:16.660 align:middle line:84%
And these credentials can
be both plaintext passwords,

00:01:16.660 --> 00:01:19.060 align:middle line:90%
but also password hashes.

00:01:19.060 --> 00:01:21.700 align:middle line:84%
What you can see on the
image currently of the side,

00:01:21.700 --> 00:01:25.100 align:middle line:84%
we can see that it would
display the username,

00:01:25.100 --> 00:01:32.200 align:middle line:84%
which is Gentil Kiwi, and also
the password waza1234/ This is

00:01:32.200 --> 00:01:33.760 align:middle line:90%
the plaintext password.

00:01:33.760 --> 00:01:36.340 align:middle line:84%
But what we can also
see, just above that,

00:01:36.340 --> 00:01:40.420 align:middle line:84%
is that it will also contain
the NTLM password hash.

00:01:40.420 --> 00:01:43.030 align:middle line:84%
While this is correct
in some occasions,

00:01:43.030 --> 00:01:48.160 align:middle line:84%
in newer systems, which are
Windows 10 and Server 2016,

00:01:48.160 --> 00:01:51.820 align:middle line:84%
we should not be seeing new, or
at least plaintext credentials

00:01:51.820 --> 00:01:52.960 align:middle line:90%
being stored.

00:01:52.960 --> 00:01:54.700 align:middle line:84%
But unfortunately,
that is sort of being

00:01:54.700 --> 00:01:56.680 align:middle line:90%
controlled by a registry key.

00:01:56.680 --> 00:01:59.850 align:middle line:84%
So an attacker who had
gained administrative access

00:01:59.850 --> 00:02:02.440 align:middle line:84%
to that endpoint, may
be able to just flip

00:02:02.440 --> 00:02:04.930 align:middle line:84%
the value in that register
key and start caching

00:02:04.930 --> 00:02:06.610 align:middle line:90%
credentials in plaintext.

00:02:06.610 --> 00:02:09.430 align:middle line:84%
But, to be honest,
the truth being told,

00:02:09.430 --> 00:02:11.560 align:middle line:84%
it doesn't matter if the
credential is in plaintext

00:02:11.560 --> 00:02:12.220 align:middle line:90%
or not.

00:02:12.220 --> 00:02:15.760 align:middle line:84%
The password hash will be more
than sufficient to actually

00:02:15.760 --> 00:02:18.010 align:middle line:84%
utilize it to remotely
connect to other systems,

00:02:18.010 --> 00:02:21.800 align:middle line:84%
unless other [? preventures ?]
are in place.

00:02:21.800 --> 00:02:23.590 align:middle line:84%
So what we're going
to demonstrate now

00:02:23.590 --> 00:02:25.720 align:middle line:84%
is, we will perform
credential theft

00:02:25.720 --> 00:02:29.120 align:middle line:84%
and using that new identity
in our environment.

00:02:29.120 --> 00:02:31.970 align:middle line:90%
So let's just switch to that.

00:02:31.970 --> 00:02:34.070 align:middle line:84%
Now, I am connected
to the environment

00:02:34.070 --> 00:02:37.010 align:middle line:84%
where we are going to
showcase the credential theft.

00:02:37.010 --> 00:02:39.380 align:middle line:84%
I have already downloaded
two tools which you need,

00:02:39.380 --> 00:02:42.260 align:middle line:90%
which are Mimikatz and Rubeus.

00:02:42.260 --> 00:02:44.990 align:middle line:84%
And to set the
scenes, let's assume

00:02:44.990 --> 00:02:48.200 align:middle line:84%
that we are an attacker
who has compromised

00:02:48.200 --> 00:02:50.760 align:middle line:90%
a user in our environment.

00:02:50.760 --> 00:02:52.940 align:middle line:84%
And one of the perks
which we will just

00:02:52.940 --> 00:02:55.010 align:middle line:84%
go forward with and
assume that we have,

00:02:55.010 --> 00:02:58.490 align:middle line:84%
is that the account or the
user which we have compromised,

00:02:58.490 --> 00:03:01.130 align:middle line:90%
has local admin rights.

00:03:01.130 --> 00:03:06.100 align:middle line:84%
Now, let's go into Mimikatz
and start with this tool.

00:03:06.100 --> 00:03:08.020 align:middle line:84%
I'm going to go
forward and start

00:03:08.020 --> 00:03:11.310 align:middle line:90%
PowerShell as an administrator.

00:03:11.310 --> 00:03:14.160 align:middle line:84%
Now, the first thing that
I want to show in here,

00:03:14.160 --> 00:03:19.020 align:middle line:84%
is that my user doesn't
have any privileged access

00:03:19.020 --> 00:03:20.910 align:middle line:90%
towards the domain controller.

00:03:20.910 --> 00:03:25.380 align:middle line:84%
So if I try to list the C
drive of the domain controller,

00:03:25.380 --> 00:03:28.470 align:middle line:84%
I'll get that there's
access denied,

00:03:28.470 --> 00:03:30.660 align:middle line:84%
so I can't see
anything in there.

00:03:30.660 --> 00:03:33.510 align:middle line:90%
Next, I was just start Mimikatz.

00:03:33.510 --> 00:03:38.220 align:middle line:84%
And inside Mimikatz,
we will execute debug--

00:03:38.220 --> 00:03:44.100 align:middle line:84%
Sorry, privilege::degbug, which
will assign us debug rights

00:03:44.100 --> 00:03:48.570 align:middle line:84%
on this machine, so it can debug
the process LSASS and extract

00:03:48.570 --> 00:03:49.992 align:middle line:90%
credentials out of it.

00:03:49.992 --> 00:03:52.200 align:middle line:84%
Then the command that we
need to execute to get those

00:03:52.200 --> 00:03:56.730 align:middle line:84%
credentials out is
sekurlsa::logonpasswords.

00:03:56.730 --> 00:03:59.460 align:middle line:90%


00:03:59.460 --> 00:04:03.240 align:middle line:84%
By doing this, the tool will
go locate the LSASS process

00:04:03.240 --> 00:04:05.820 align:middle line:84%
and extract all the
credentials back to us

00:04:05.820 --> 00:04:09.990 align:middle line:84%
in a human readable,
understandable format.

00:04:09.990 --> 00:04:12.510 align:middle line:84%
And because this is
a Windows 10 machine,

00:04:12.510 --> 00:04:15.540 align:middle line:84%
we don't expect to see
any plaintext passwords.

00:04:15.540 --> 00:04:17.610 align:middle line:84%
And what we are looking
at in this output

00:04:17.610 --> 00:04:21.420 align:middle line:84%
is any user or
credentials of something

00:04:21.420 --> 00:04:23.310 align:middle line:90%
that is more privileged.

00:04:23.310 --> 00:04:28.080 align:middle line:84%
So what we can see in here is
that the account administrator

00:04:28.080 --> 00:04:33.150 align:middle line:84%
must have logged in, but nothing
is stored from that session.

00:04:33.150 --> 00:04:35.220 align:middle line:84%
Then we will continue
scrolling up.

00:04:35.220 --> 00:04:37.710 align:middle line:84%
We can see that there
is a login from Slavi.

00:04:37.710 --> 00:04:41.990 align:middle line:84%
In this case, we will see
the NTLM password hash.

00:04:41.990 --> 00:04:46.740 align:middle line:84%
And continue scrolling
through the output.

00:04:46.740 --> 00:04:48.480 align:middle line:90%
Going up, up.

00:04:48.480 --> 00:04:50.770 align:middle line:84%
And actually, if
you look in here.

00:04:50.770 --> 00:04:54.150 align:middle line:84%
Now we will see that
there is the administrator

00:04:54.150 --> 00:04:57.000 align:middle line:84%
and the password hash
is also available.

00:04:57.000 --> 00:04:59.010 align:middle line:84%
Now, this is interesting,
because this is really

00:04:59.010 --> 00:05:00.540 align:middle line:90%
what we want to achieve.

00:05:00.540 --> 00:05:03.612 align:middle line:84%
Actually, extract
these credentials.

00:05:03.612 --> 00:05:05.320 align:middle line:84%
So essentially, we
have the password hash

00:05:05.320 --> 00:05:09.270 align:middle line:84%
for the account administrator
of the domain PROD.

00:05:09.270 --> 00:05:13.590 align:middle line:84%
Now, it's time to use our second
tool, which we have downloaded,

00:05:13.590 --> 00:05:15.240 align:middle line:90%
and that is Rubeus.

00:05:15.240 --> 00:05:18.520 align:middle line:84%
I will just start Rubeus
in a separate window,

00:05:18.520 --> 00:05:21.750 align:middle line:84%
so we can keep track of
which window is doing what.

00:05:21.750 --> 00:05:28.440 align:middle line:84%
Now, in Rubeus, I will specify
the parameter asktgt, which

00:05:28.440 --> 00:05:30.960 align:middle line:90%
will create a Kerberos ticket.

00:05:30.960 --> 00:05:34.410 align:middle line:84%
But now, this ticket will
also specify the username

00:05:34.410 --> 00:05:36.640 align:middle line:84%
that we want the ticket
to be created for,

00:05:36.640 --> 00:05:39.150 align:middle line:90%
and it is administrator.

00:05:39.150 --> 00:05:43.260 align:middle line:84%
And we are going to specify
the NTLM password hash

00:05:43.260 --> 00:05:46.210 align:middle line:90%
of this account administrator.

00:05:46.210 --> 00:05:49.780 align:middle line:84%
So Rubeus will obtain
a Kerberos ticket

00:05:49.780 --> 00:05:53.230 align:middle line:84%
for the account
administrator by just

00:05:53.230 --> 00:05:55.450 align:middle line:84%
utilizing this special
hash, without knowing

00:05:55.450 --> 00:05:57.160 align:middle line:90%
the plaintext password.

00:05:57.160 --> 00:05:59.140 align:middle line:84%
And the next parameter
that I want to provide

00:05:59.140 --> 00:06:02.170 align:middle line:84%
is PTT, which is
passing the ticket.

00:06:02.170 --> 00:06:05.080 align:middle line:84%
The Kerberos ticket that
Rubeus is going to extract,

00:06:05.080 --> 00:06:06.790 align:middle line:84%
it would just
automatically inject it

00:06:06.790 --> 00:06:08.280 align:middle line:90%
into this PowerShell session.

00:06:08.280 --> 00:06:10.745 align:middle line:84%
So in this session, I'll be
running as the administrator

00:06:10.745 --> 00:06:13.740 align:middle line:90%
account, if this succeeds.

00:06:13.740 --> 00:06:16.710 align:middle line:84%
So running the command
will get the actual ticket,

00:06:16.710 --> 00:06:18.780 align:middle line:90%
and it's being injected.

00:06:18.780 --> 00:06:22.380 align:middle line:84%
It's notified by this
message here that the ticket

00:06:22.380 --> 00:06:24.120 align:middle line:90%
was successfully imported.

00:06:24.120 --> 00:06:27.930 align:middle line:84%
If I do klist, I will
see that there is indeed

00:06:27.930 --> 00:06:31.500 align:middle line:84%
a Kerberos ticket in
my PowerShell session

00:06:31.500 --> 00:06:36.640 align:middle line:84%
for the account administrator
in this production domain.

00:06:36.640 --> 00:06:39.420 align:middle line:84%
So if I perform the
initial step again,

00:06:39.420 --> 00:06:42.782 align:middle line:84%
by trying to execute the
ls command against the C

00:06:42.782 --> 00:06:44.760 align:middle line:84%
shared drive of
domain controller,

00:06:44.760 --> 00:06:48.360 align:middle line:84%
now we will actually
see the content of it.

00:06:48.360 --> 00:06:50.730 align:middle line:84%
We are, essentially, the
administrator account.

00:06:50.730 --> 00:06:53.190 align:middle line:84%
We have successfully
compromised this account.

00:06:53.190 --> 00:06:54.810 align:middle line:84%
And what is crazy
is that we never

00:06:54.810 --> 00:06:58.500 align:middle line:84%
even knew the password of it,
we only had the password hash.

00:06:58.500 --> 00:07:00.810 align:middle line:84%
And this is the
credential theft that we

00:07:00.810 --> 00:07:02.610 align:middle line:90%
want to really prevent against.

00:07:02.610 --> 00:07:05.610 align:middle line:84%
And one of the main reasons
why this is possible,

00:07:05.610 --> 00:07:09.000 align:middle line:84%
is because privileged users,
such as this administrator

00:07:09.000 --> 00:07:15.420 align:middle line:84%
account, are allowed to log into
non-secure or often compromised

00:07:15.420 --> 00:07:19.730 align:middle line:84%
devices, such as
endpoint workstations.

00:07:19.730 --> 00:07:21.890 align:middle line:84%
Now, let's go back to
the slides and talk

00:07:21.890 --> 00:07:24.440 align:middle line:84%
about the model which will
try to prevent this credential

00:07:24.440 --> 00:07:25.250 align:middle line:90%
theft.

00:07:25.250 --> 00:07:27.140 align:middle line:84%
The Active Directory
Tiering model

00:07:27.140 --> 00:07:29.150 align:middle line:84%
is designed to
provide protection

00:07:29.150 --> 00:07:32.030 align:middle line:84%
of stealing highly privileged
credentials by implementing

00:07:32.030 --> 00:07:33.140 align:middle line:90%
buffer zones.

00:07:33.140 --> 00:07:37.190 align:middle line:84%
In other words, these
buffer zones or tiers

00:07:37.190 --> 00:07:40.220 align:middle line:84%
are categorizations of
assault. By creating

00:07:40.220 --> 00:07:42.530 align:middle line:84%
this categorization,
we will enforce

00:07:42.530 --> 00:07:45.050 align:middle line:84%
an amount of restrictions
in our organization

00:07:45.050 --> 00:07:48.020 align:middle line:84%
that will prevent privileged
users from accessing

00:07:48.020 --> 00:07:51.500 align:middle line:84%
non-privileged, or rather,
less trusted devices.

00:07:51.500 --> 00:07:54.110 align:middle line:84%
By doing so, we will
effectively mitigate

00:07:54.110 --> 00:07:57.270 align:middle line:84%
the risk of low privileged
users from performing

00:07:57.270 --> 00:08:00.870 align:middle line:84%
credential theft against
a highly privileged user.

00:08:00.870 --> 00:08:03.390 align:middle line:84%
The default model
introduced by Microsoft,

00:08:03.390 --> 00:08:08.340 align:middle line:84%
contains three tiers, which
are tier two, representing

00:08:08.340 --> 00:08:10.570 align:middle line:90%
workstations and their users.

00:08:10.570 --> 00:08:12.960 align:middle line:84%
It also includes
groups at this level.

00:08:12.960 --> 00:08:16.320 align:middle line:84%
Both for regular users, but
also administrative groups

00:08:16.320 --> 00:08:20.400 align:middle line:84%
for those users who perform
administration on workstations,

00:08:20.400 --> 00:08:22.560 align:middle line:90%
such as help desk personnel.

00:08:22.560 --> 00:08:25.310 align:middle line:84%
Tier two represents
regular member servers

00:08:25.310 --> 00:08:28.290 align:middle line:84%
in the domain and
their associated users,

00:08:28.290 --> 00:08:31.230 align:middle line:84%
groups and applications
running on them.

00:08:31.230 --> 00:08:35.340 align:middle line:84%
Tier zero is defined as the
most critical infrastructure,

00:08:35.340 --> 00:08:38.820 align:middle line:84%
such as domain controllers,
PKI, and so on.

00:08:38.820 --> 00:08:41.039 align:middle line:84%
It, again, includes
the users endgroups

00:08:41.039 --> 00:08:44.430 align:middle line:84%
who operate at this
level, such as domain

00:08:44.430 --> 00:08:46.410 align:middle line:90%
administrator and similar.

00:08:46.410 --> 00:08:49.200 align:middle line:84%
In this next slide,
we are representing

00:08:49.200 --> 00:08:53.820 align:middle line:84%
the different actions that are
allowed to occur in our tiers.

00:08:53.820 --> 00:08:57.570 align:middle line:84%
Tier two administrators can
only manage and control assets

00:08:57.570 --> 00:08:59.310 align:middle line:90%
at the tier two level.

00:08:59.310 --> 00:09:01.680 align:middle line:84%
These users are
forbidden from accessing

00:09:01.680 --> 00:09:03.910 align:middle line:90%
any of the other tiers.

00:09:03.910 --> 00:09:07.050 align:middle line:84%
Tier one administrators can
manage and control assets

00:09:07.050 --> 00:09:10.940 align:middle line:90%
at tier one and tier two levels.

00:09:10.940 --> 00:09:13.310 align:middle line:84%
But as we'll see
later, these users

00:09:13.310 --> 00:09:17.160 align:middle line:84%
can only login interactively
to tier one systems.

00:09:17.160 --> 00:09:20.180 align:middle line:84%
Finally, tier zero can
manage and control assets

00:09:20.180 --> 00:09:21.510 align:middle line:90%
at any level.

00:09:21.510 --> 00:09:24.990 align:middle line:84%
An example I often give,
is to think of GPOs.

00:09:24.990 --> 00:09:27.230 align:middle line:84%
If you are on a
domain controller,

00:09:27.230 --> 00:09:29.780 align:middle line:84%
you can create or
modify GPOs that

00:09:29.780 --> 00:09:32.120 align:middle line:84%
can be applied to
devices on all tiers.

00:09:32.120 --> 00:09:35.000 align:middle line:90%


00:09:35.000 --> 00:09:37.370 align:middle line:84%
Now that we have discussed
the control part,

00:09:37.370 --> 00:09:39.660 align:middle line:84%
let's take a look
at allowed logins.

00:09:39.660 --> 00:09:42.410 align:middle line:84%
I suspect that there's
no surprises here.

00:09:42.410 --> 00:09:47.420 align:middle line:84%
In short, a lower tier is not
allowed to access higher tier.

00:09:47.420 --> 00:09:50.690 align:middle line:84%
For example, a tier
zero count is not

00:09:50.690 --> 00:09:55.586 align:middle line:84%
allowed to access a
tier two workstation.

00:09:55.586 --> 00:09:58.850 align:middle line:84%
And this is really what
the entire model is about.

00:09:58.850 --> 00:10:02.450 align:middle line:84%
If privileged users cannot
access untrusted systems,

00:10:02.450 --> 00:10:04.670 align:middle line:84%
then there will be no
credentials to be stolen from

00:10:04.670 --> 00:10:06.340 align:middle line:90%
them.

00:10:06.340 --> 00:10:08.650 align:middle line:84%
For this model to
be effective, I

00:10:08.650 --> 00:10:11.470 align:middle line:84%
have listed some
requirements that you really

00:10:11.470 --> 00:10:13.300 align:middle line:90%
should consider implementing.

00:10:13.300 --> 00:10:16.120 align:middle line:84%
The first one is rather
obvious, that servers

00:10:16.120 --> 00:10:20.740 align:middle line:84%
should be domain
joined, otherwise

00:10:20.740 --> 00:10:24.700 align:middle line:84%
we can't really talk about
a tiering model of a domain.

00:10:24.700 --> 00:10:27.560 align:middle line:84%
Then I recommend that
administrative users,

00:10:27.560 --> 00:10:31.120 align:middle line:84%
especially those at tier zero,
are members of the Protected

00:10:31.120 --> 00:10:34.240 align:middle line:84%
Users group, to benefit
from all of the protections

00:10:34.240 --> 00:10:36.100 align:middle line:90%
that this group provides.

00:10:36.100 --> 00:10:38.200 align:middle line:84%
Of course, unfortunately,
they will also

00:10:38.200 --> 00:10:40.570 align:middle line:84%
inherit all the
problematic perks

00:10:40.570 --> 00:10:42.590 align:middle line:84%
that this membership
brings with it.

00:10:42.590 --> 00:10:45.640 align:middle line:84%
So extensive testing before
adding users in this group

00:10:45.640 --> 00:10:49.420 align:middle line:84%
is definitely required to verify
that no critical systems are

00:10:49.420 --> 00:10:53.410 align:middle line:84%
breaking or access to those
system is being blocked.

00:10:53.410 --> 00:10:55.420 align:middle line:90%
Next on the list is LAPS.

00:10:55.420 --> 00:10:59.710 align:middle line:84%
And this is here because,
if LAPS is not installed,

00:10:59.710 --> 00:11:03.400 align:middle line:84%
then there may be
tier one servers

00:11:03.400 --> 00:11:05.810 align:middle line:84%
who have the password of
the local administrator

00:11:05.810 --> 00:11:09.670 align:middle line:84%
account being the same as the
password of the same account

00:11:09.670 --> 00:11:11.880 align:middle line:90%
in a tier zero server.

00:11:11.880 --> 00:11:14.340 align:middle line:84%
That would directly breach
and break the model,

00:11:14.340 --> 00:11:17.770 align:middle line:84%
as compromising tier one
server would provide access

00:11:17.770 --> 00:11:19.990 align:middle line:90%
to tier zero systems.

00:11:19.990 --> 00:11:22.220 align:middle line:84%
Patching is also
really important.

00:11:22.220 --> 00:11:25.480 align:middle line:84%
If we don't patch our
systems, then adversaries

00:11:25.480 --> 00:11:28.660 align:middle line:84%
may have the opportunity to
compromise tier one or tier

00:11:28.660 --> 00:11:32.890 align:middle line:84%
zero unpatched servers from
a tier two endpoint device.

00:11:32.890 --> 00:11:34.930 align:middle line:84%
Of course, we do not
want this to happen.

00:11:34.930 --> 00:11:38.320 align:middle line:84%
So effective patch management
should be in place.

00:11:38.320 --> 00:11:42.560 align:middle line:84%
And note that the recording
of this is in 2020,

00:11:42.560 --> 00:11:45.920 align:middle line:84%
and we are still talking about
patching being a problem.

00:11:45.920 --> 00:11:49.240 align:middle line:84%
So let's hope that the future
will bring better days.

00:11:49.240 --> 00:11:52.930 align:middle line:84%
The final bullet I have outlined
is workstation and server

00:11:52.930 --> 00:11:53.830 align:middle line:90%
hardening.

00:11:53.830 --> 00:11:57.040 align:middle line:84%
I can't recommend enough to
implement better protection

00:11:57.040 --> 00:12:00.640 align:middle line:84%
to endpoint devices by really
limiting the [? attacked ?]

00:12:00.640 --> 00:12:01.700 align:middle line:90%
servers to them.

00:12:01.700 --> 00:12:04.060 align:middle line:84%
You should really
implement some features

00:12:04.060 --> 00:12:07.060 align:middle line:84%
such as Credential Guard,
AppLocker, BitLocker,

00:12:07.060 --> 00:12:09.340 align:middle line:84%
the Windows Firewall,
especially for blocking

00:12:09.340 --> 00:12:12.370 align:middle line:84%
incoming connections,
and also restricting

00:12:12.370 --> 00:12:16.000 align:middle line:84%
the execution of PowerShell
for regular users.

00:12:16.000 --> 00:12:18.820 align:middle line:84%
Something additional
on managing systems

00:12:18.820 --> 00:12:21.940 align:middle line:84%
is the usage of the principle
of least privileged.

00:12:21.940 --> 00:12:24.700 align:middle line:84%
Having as little as
possible administrator,

00:12:24.700 --> 00:12:27.820 align:middle line:84%
of about two or
three, is key driver

00:12:27.820 --> 00:12:30.340 align:middle line:84%
to reducing lateral
movement in the environment.

00:12:30.340 --> 00:12:33.380 align:middle line:84%
Also, privileged delegation
all over the place

00:12:33.380 --> 00:12:35.710 align:middle line:84%
is really not necessary,
especially when

00:12:35.710 --> 00:12:37.210 align:middle line:90%
it's not controlled.

00:12:37.210 --> 00:12:39.730 align:middle line:84%
Tools, such as Privileged
Access Management,

00:12:39.730 --> 00:12:43.480 align:middle line:84%
are really good at this,
because escalated access can

00:12:43.480 --> 00:12:45.640 align:middle line:84%
be provisioned only
for a short time

00:12:45.640 --> 00:12:49.000 align:middle line:84%
and automatically
removed afterwards.

00:12:49.000 --> 00:12:52.240 align:middle line:84%
Also, privileged delegation
all over the place

00:12:52.240 --> 00:12:56.140 align:middle line:84%
is not necessary, especially
when it's uncontrolled.

00:12:56.140 --> 00:12:58.310 align:middle line:84%
Tools, such as Privileged
Access Management,

00:12:58.310 --> 00:13:02.110 align:middle line:84%
are really good at this,
because escalated access can

00:13:02.110 --> 00:13:04.370 align:middle line:84%
be provisioned only
for a short time

00:13:04.370 --> 00:13:08.040 align:middle line:84%
and then automatically
removed afterwards.

00:13:08.040 --> 00:13:10.830 align:middle line:84%
On the next slide, I have
a table taken directly

00:13:10.830 --> 00:13:14.220 align:middle line:84%
from Microsoft, which is
found on the link that

00:13:14.220 --> 00:13:18.120 align:middle line:84%
is at the bottom of the slide,
in which Microsoft showcases

00:13:18.120 --> 00:13:22.280 align:middle line:84%
their recommendation
on operation practices.

00:13:22.280 --> 00:13:24.950 align:middle line:84%
These are all very
relevant and should be

00:13:24.950 --> 00:13:26.780 align:middle line:90%
considered to be implemented.

00:13:26.780 --> 00:13:29.030 align:middle line:84%
What I wanted to
outline especially,

00:13:29.030 --> 00:13:33.170 align:middle line:84%
is the Manage Risk of
Credential Exposed section.

00:13:33.170 --> 00:13:35.330 align:middle line:84%
While you can see that what
we discussed previously

00:13:35.330 --> 00:13:38.300 align:middle line:84%
on least privileged is
also mentioned here,

00:13:38.300 --> 00:13:42.500 align:middle line:84%
and Microsoft also flags
as mandatory that tier zero

00:13:42.500 --> 00:13:45.740 align:middle line:84%
accounts should use a
strong authentication,

00:13:45.740 --> 00:13:48.200 align:middle line:84%
such as smartcard
token, or other type

00:13:48.200 --> 00:13:50.210 align:middle line:84%
of multi-factor
indication will also

00:13:50.210 --> 00:13:54.050 align:middle line:84%
suffice under the assumption
that administrators have

00:13:54.050 --> 00:13:56.490 align:middle line:90%
long and complex passwords.

00:13:56.490 --> 00:13:58.850 align:middle line:84%
Now, for the credential
exposure part,

00:13:58.850 --> 00:14:01.850 align:middle line:84%
we need to have a separate
administrative accounts.

00:14:01.850 --> 00:14:04.820 align:middle line:84%
General Internet browsing
and reading emails

00:14:04.820 --> 00:14:08.150 align:middle line:84%
is not allowed on
administrative workstations.

00:14:08.150 --> 00:14:10.490 align:middle line:84%
These workstation
that I'm referring to

00:14:10.490 --> 00:14:13.640 align:middle line:84%
is known as privileged
access workstation,

00:14:13.640 --> 00:14:16.860 align:middle line:84%
and we will discuss this in
detail at the later model.

00:14:16.860 --> 00:14:20.300 align:middle line:84%
But for now, just think
of it as a separate device

00:14:20.300 --> 00:14:23.480 align:middle line:84%
that administrators
use to access tier zero

00:14:23.480 --> 00:14:26.490 align:middle line:90%
and tier one management systems.

00:14:26.490 --> 00:14:30.390 align:middle line:84%
There is also a mention of
administrator logon practices.

00:14:30.390 --> 00:14:32.610 align:middle line:84%
And to describe this
in better detail,

00:14:32.610 --> 00:14:36.780 align:middle line:84%
I created the following table
to outline allowed and forbidden

00:14:36.780 --> 00:14:37.410 align:middle line:90%
logins.

00:14:37.410 --> 00:14:39.270 align:middle line:84%
In the table, the
different categories

00:14:39.270 --> 00:14:41.790 align:middle line:84%
which I have provided
are practices

00:14:41.790 --> 00:14:44.790 align:middle line:84%
for remote logins,
for physical logins,

00:14:44.790 --> 00:14:48.540 align:middle line:84%
and what should not occur
across the different tiers.

00:14:48.540 --> 00:14:51.810 align:middle line:84%
Let's take, as an
example, just tier two.

00:14:51.810 --> 00:14:54.240 align:middle line:84%
In here, we allow
Remote Assistance

00:14:54.240 --> 00:14:57.450 align:middle line:84%
over Skype or other
screen sharing tools,

00:14:57.450 --> 00:15:02.380 align:middle line:84%
as well as RDP, only when
Remote Credential Guard is used.

00:15:02.380 --> 00:15:05.880 align:middle line:84%
We also allow administrative
access to the workstation

00:15:05.880 --> 00:15:09.420 align:middle line:84%
by using the administrator
account managed by LAPS.

00:15:09.420 --> 00:15:13.560 align:middle line:84%
In many cases, this will be the
built-in administrator account.

00:15:13.560 --> 00:15:18.310 align:middle line:84%
For physical logins, we allow
only usage of LAPS account.

00:15:18.310 --> 00:15:22.100 align:middle line:84%
And finally, we forbid login
with administrative domain

00:15:22.100 --> 00:15:26.430 align:middle line:84%
account, and we do not perform
any normal RDP connection

00:15:26.430 --> 00:15:30.660 align:middle line:84%
with a privileged domain account
towards these endpoint devices.

00:15:30.660 --> 00:15:32.850 align:middle line:84%
By following this
table, we are really

00:15:32.850 --> 00:15:36.030 align:middle line:84%
limiting the exposed credentials
that will, at any point,

00:15:36.030 --> 00:15:38.170 align:middle line:90%
reside at an endpoint.

00:15:38.170 --> 00:15:40.470 align:middle line:84%
It may seem that
this model sounds

00:15:40.470 --> 00:15:43.350 align:middle line:84%
complex and hard to
follow, but trust me,

00:15:43.350 --> 00:15:45.970 align:middle line:84%
once you understand
it, it's very easy.

00:15:45.970 --> 00:15:47.530 align:middle line:90%
And in this video--

00:15:47.530 --> 00:15:48.810 align:middle line:90%
I'll stop it here--

00:15:48.810 --> 00:15:50.670 align:middle line:84%
we are wrapping up
the theoretical part

00:15:50.670 --> 00:15:52.210 align:middle line:90%
behind this tiering model.

00:15:52.210 --> 00:15:54.060 align:middle line:84%
While in the next
video, we will look

00:15:54.060 --> 00:15:56.970 align:middle line:84%
into the practical
implementation of this model

00:15:56.970 --> 00:16:00.540 align:middle line:84%
and how we can actually achieve
these forbidden and allowed

00:16:00.540 --> 00:16:02.090 align:middle line:90%
logins.

00:16:02.090 --> 00:16:03.000 align:middle line:90%