NewAI Security Practice — securing the AI systems your business now depends on.

BBPen · Other

Advanced Black Box Penetration Testing

Care for a game of capture-the-flag? A goal-based, adversary-style engagement with no internal knowledge, no whitelisting, and minimal scope restrictions. The closest thing to a real attack short of one.

When PSA isn't enough

A standard penetration test asks: which vulnerabilities exist on this scope? A black-box engagement asks: can a determined adversary achieve this objective? The two answer different questions. BBPen is for organizations that already do regular PSA-style testing and need to know whether their actual defensive program — people, process, technology, detection, response — works against an attacker who refuses to follow the script.

Boards and audit committees increasingly ask a single question: "if a serious adversary targeted us, would we know? Would we stop them?" BBPen is how you find out before they do.

The kill chain

1

Goal definition

We define the 'flags' with you in advance: domain admin? PII exfiltration? Wire transfer? Source code? Production database? The goal drives everything that follows.

2

OSINT & reconnaissance

Public footprint, employee enumeration, exposed credentials, third-party trust relationships, physical site scouting. Same playbook as a real adversary.

3

Initial access

Multi-vector — perimeter exploitation, phishing & vishing, malicious USB drops, physical pretext, supplier compromise. Whichever vector works first wins.

4

Foothold & evasion

Establish C2 with techniques designed to evade your EDR, SIEM, and SOC. We measure detection and response as a core deliverable.

5

Lateral movement

Credential harvesting, AD attack paths, Kerberoasting, ticket abuse, privilege escalation across cloud and on-prem boundaries.

6

Objective & exfiltration

Reach the flag. Demonstrate impact safely. Exfiltrate using techniques real attackers use — DNS tunneling, cloud sync abuse, encrypted channels.

7

Debrief & remediation

Full timeline, the kill chain that worked, the controls that fired (or didn't), and prioritized remediation tied to your detection gaps.

Attack vectors

Network

Perimeter exploitation, exposed services, VPN abuse, supply-chain pivots.

Web & API

Auth bypass, SSRF to cloud metadata, IDOR chains to admin functions.

Social

Spear-phishing, vishing, pretext-based on-site engagements, malicious USB.

Wireless

Evil-twin AP, WPA2/3-Enterprise relay, BLE attacks against badge systems.

Physical

Tailgating, badge cloning, lock bypass, planted implants and rogue devices.

Supply chain

Compromise of trusted third parties, malicious package vectors, vendor accounts.

Rules of engagement

  • Black-box, goal-driven scoping defined with you up front
  • No internal knowledge provided to the testing team
  • No source IPs whitelisted in detection tooling
  • Defined out-of-scope assets and safety lanes
  • Designated 'trusted agents' inside your org for emergency contact only
  • Daily go/no-go check-ins with the lead tester (not your SOC)
  • Optional 'purple team' phase after the engagement to walk through what worked

Highlights

  • Black-box, goal-driven scoping
  • OSINT-led attack planning
  • Multi-vector attack chains (network, web, social, physical)
  • Stealth, evasion, and persistence techniques
  • Custom-built tooling and exploits where needed
  • Detection and response measurement
  • Full kill-chain timeline in deliverable

Ready to start a conversation?

Talk to a senior consultant — we'll scope an engagement that fits your environment.