Internal Penetration Test

A serious attack often starts inside the network: one workstation compromised through a phishing email, an employee with malicious intent (malicious insider) or a guest connected to the Wi-Fi. An internal penetration test checks what can be achieved from there, how close the attacker gets to admin rights in Active Directory, to the servers and to the backups, and what needs to be fixed to stop them along the way.

Server racks with neat cabling

From one workstation
to the heart of the network

The test starts from a point defined in advance: a network connection only, a regular user, or an employee workstation that has already been compromised (assumed breach). From there we check how far it’s possible to get, map the attack paths in Active Directory and mark the points where a single fix blocks an entire path.

  • Attack paths and excessive privileges in Active Directory
  • Weak passwords, the same password on several accounts (password reuse) and legacy protocols
  • Network separation (segmentation), lateral movement and access to backups

Call +972-3-3801180

Who it’s for

  • Organizations with Active Directory

    Whose domain was built over years and passed through several providers, and nobody has checked who actually holds admin rights (privileged accounts).

  • Offices with several branches or a flat network

    Where every workstation sees every server (flat network), and it’s unclear whether the separation between network zones (segmentation, VLAN) holds up under testing.

  • Organizations with suppliers and guests on the network

    Where a contractor, a remote support provider (third-party access) or a guest on the Wi-Fi connects to the same infrastructure that holds the sensitive data.

Sound familiar?

These are findings an internal test often turns up, even in networks that look well organized.

  • Service accounts with domain admin rights and a password that hasn’t been changed in years
  • The same local admin password on every computer, with no unique password per computer (LAPS)
  • Employees who changed roles and kept all their previous permissions (privilege creep)
  • Old servers that no longer receive security updates (end of life) and are still connected to the network
  • A guest network that is actually connected to the office network
  • The backup server is joined to the domain, and anyone who gets admin rights reaches it too

What an internal test checks

The scope and the starting point are set in advance, with written authorization and agreed rules of engagement.

  • Attack paths in Active Directory

    Mapping the relationships between users, groups, computers and permissions, to find the shortest route from a regular user to domain admin rights (Domain Admin).

  • Excessive privileges

    Membership in admin groups, permissions that were delegated and forgotten (delegation), service accounts and accounts of employees who left.

  • Passwords and reuse

    The password policy, weak or default passwords, the same password on several accounts (password reuse) and a shared local admin password.

  • Legacy protocols

    Protocols left active by default, such as SMBv1, LLMNR, NBT-NS and NTLMv1, that expose credentials on the network.

  • Unpatched servers and services

    Versions that haven’t received security updates (patch management), systems that have reached end of life and management interfaces with a default password.

  • Network separation and lateral movement

    Checking whether the separation between network zones (segmentation, VLAN, firewall rules) actually prevents moving from a workstation to the servers, the OT network and the management interfaces (lateral movement).

  • Access to backups

    Whether someone who gained permissions on the network can reach the backup server and the copies, and whether there is a copy that can’t be changed or deleted (immutable backup) or a disconnected copy (air-gapped).

  • Report, remediation and retest

    The attack paths found, a CVSS rating for each finding, a fix order (remediation) that blocks the short paths first, and a retest.

What we take care of, in detail

Active Directory and identities

  • Active Directory security assessment using standard tools such as PingCastle, Purple Knight and BloodHound
  • Mapping attack paths to Domain Admin and Enterprise Admin rights
  • Admin groups, delegated permissions (delegation) and unusual permissions on objects (ACL)
  • Service accounts, accounts whose password never expires and inactive accounts
  • Password policy, weak passwords and password reuse between regular accounts and admin accounts
  • A unique local admin password for each computer (LAPS), separation of admin accounts (admin tiering) and the Protected Users group
  • GPO, Kerberos and certificate services (AD CS) settings that allow privilege escalation
  • The connection to Entra ID and the effect of on-premises accounts on the cloud (hybrid identity)

Network, servers and workstations

  • Different starting points: an employee workstation, a free network port, the guest network (guest Wi-Fi) and a VPN connection
  • Legacy protocols: SMBv1, LLMNR, NBT-NS, NTLMv1, and the SMB signing status
  • Unpatched servers and workstations (patch management), End of Life systems and unnecessary services
  • File shares with broad permissions and documents that contain passwords
  • Management interfaces of printers, switches, NAS, cameras and iLO / iDRAC with default passwords
  • Testing network separation (segmentation): moving between VLANs, firewall rules between zones, separation of OT and IoT
  • Lateral movement between workstations and servers, and whether endpoint protection (EDR) detects it and alerts
  • Access to backup servers, backup repositories and virtualization consoles (vCenter, Hyper-V)

Methodology and safety

  • Written authorization, scope definition and agreed rules of engagement before work starts
  • Black, grey or white box, and a scenario of a workstation that has already been compromised (assumed breach), depending on the goal of the test
  • Non-destructive methods in the production environment, avoiding account lockouts and load on critical servers
  • A coordinated testing window, an available contact person in the organization and a stop condition
  • Verifying a working backup and running a restore test before the test
  • Updating the contact person during the test as soon as a critical finding comes up

From the first conversation to an organized routine

  1. 1

    Introductory call and mapping

    The structure of the network, the domain, the branches and the sensitive systems, and the goal of the test.

  2. 2

    Scope, authorization and rules of engagement

    Starting point, time windows, a contact person and what is out of scope, with written authorization and agreed rules of engagement.

  3. 3

    Controlled execution

    Testing with methods that don’t harm the production environment, after verifying a working backup, and updating the contact person during the test about any critical finding.

  4. 4

    Report, remediation and retest

    The attack paths and the remediation recommendations, carrying out the fixes in the infrastructure, and a retest.

How an attack spreads inside the network, and where to stop it

In ransomware incidents, the major damage is usually done after the attacker gets in: at the stage where they gain permissions, move between computers and reach the backups. These are the links an internal test checks:

  1. Why Active Directory is at the center

    Active Directory determines who can access every server, folder and computer in the organization, so whoever gains admin rights in it effectively controls the entire network. A domain built over years accumulates groups, permissions and settings that nobody remembers adding, or why. Tools like PingCastle and Purple Knight give a snapshot of the settings, and BloodHound maps the attack paths between a regular user and the domain admin.

  2. Excessive permissions and forgotten accounts

    An employee who changed roles often keeps the previous permissions (privilege creep), and a service account that was once set up with admin rights keeps holding them. Each of these is a shortcut for an attacker. The principle that guides the fix is least privilege: every account gets only what its work requires.

  3. Passwords and legacy protocols

    A password repeated across several accounts (password reuse), or an identical local admin password on every computer, turns a breach of one workstation into access to all of them. Legacy protocols such as LLMNR and NTLMv1 stay active by default and expose credentials on the network. Turning them off, together with a unique password for each computer (LAPS), is among the fixes that give a lot of protection for relatively little work.

  4. Separation that exists only on the diagram

    Often the network diagram shows separate VLANs for servers, workstations and guests, while in practice the firewall rules between them are almost completely open. The test checks the separation as it actually works: whether an employee workstation can reach the management interfaces, the backup server or the OT network. Real separation (segmentation and micro-segmentation) limits lateral movement even when something has already been compromised.

  5. The backup as a target

    Attackers look for the backups before they launch ransomware, because an available backup lets the organization restore without paying. A backup server joined to the domain, with the same admin passwords, falls together with the rest of the network. The test checks whether the backup is separated, whether there is a copy that can’t be changed or deleted (immutable backup) and whether there is a disconnected copy (air-gapped).

Guide: Active Directory: why it is an attacker’s first target, and how to assess it

Frequently asked questions

About internal penetration testing.

Didn’t find an answer? +972-3-3801180

The firewall protects the boundary between the network and the internet (the perimeter). An attack that starts with a phishing email, a leaked password or a supplier’s workstation is already inside, and from there the question is what the internal settings allow. An internal penetration test checks exactly this stage, which decides whether the incident stops at one workstation or spreads across the whole organization.

From a starting point defined in advance according to the scenario you want to test: a network port without a user, a regular employee user (assumed breach), a specific workstation or the guest network. Sometimes several starting points are tested in the same engagement. The choice is made when the scope is defined and is written into the rules of engagement.

The test is planned to minimize that risk: we check the account lockout policy before any password testing, use non-destructive methods and coordinate a time window in advance. Before work starts we make sure there is a working backup, and a contact person in the organization is available throughout the test. If something behaves unusually, we stop and update them.

PingCastle and Purple Knight scan the Active Directory settings and give an important snapshot, and they are used as part of the test. A penetration test connects the findings: it checks whether one weak setting, together with a shared password and an unpatched server, creates a path from a regular user to admin rights. It also tests the network, the servers and the backups, which these tools don’t cover.

A report with an executive summary and a description of the attack paths found, findings rated by CVSS and a remediation recommendation for each one. The fix order starts with the points that block the shortest paths to admin rights. After the fixes we carry out a retest and update the report.

When we manage the infrastructure, we do the fixes: permissions, GPO, LAPS, disabling protocols, segmentation and separating the backup, coordinated with work needs and with change documentation (change management). When the infrastructure is managed by another provider, we hand the findings over in an orderly way, support the fix and test again.

Want to know how far an attacker would get from one workstation on your network?

Tell us briefly about your business and what concerns you about your IT. We’ll get back to you to understand what you need and suggest a way of working that fits.

Business hours: Sunday to Thursday, 08:00 to 17:00

Existing client? Open a support ticket

Leave your details and we’ll get back to you

Your details are used only to get back to you. Privacy policy

Call +972-3-3801180Get in touch