Cybernion
← Insights

Security

Types of Penetration Testing Explained

30 Aug 2026 · Updated 30 Aug 2026

Penetration tests are classified three ways: by how much the tester knows (black, grey or white box), by the tester’s position and whether staff are aware (external or internal, overt or covert), and by target (network, web, API, mobile, wireless, cloud, social engineering, physical). Understanding the types is what lets an organisation commission the test that actually matches its risk, rather than a generic one.

“Penetration test” covers a wide range of engagements, and buying the wrong type means paying for assurance that was not needed while missing the assurance that was. Here is how the types break down.

What are the main types of penetration testing?

Penetration tests are classified along three independent dimensions, and a real engagement usually combines them. The first is how much the tester knows about the target. The second is the tester’s position and whether the organisation’s staff know the test is happening. The third is what is being tested, the target.

So a single engagement might be, for example, a grey box, external, web application test, or a covert, internal, network test. None of the three dimensions replaces the others. Getting the combination right is what makes a test useful: the knowledge level sets how realistic and how thorough it is, the position sets whether the test looks at the outside or the inside and whether detection is in scope too, and the target sets what actually gets examined. The rest of this article takes each dimension in turn, and the underlying methodologies are covered in penetration testing methodology.

Black box, grey box and white box testing

This dimension is about how much information the tester is given. The US standards body NIST defines the three levels precisely. Black box testing assumes no knowledge of the internal structure and implementation detail of the target, so the tester works with nothing but what an outsider could discover. White box testing assumes explicit and substantial knowledge, including internal detail such as architecture and source code. Grey box testing assumes some knowledge, for example user level credentials or limited documentation.

Each has a trade-off. Black box most closely mimics a real external attacker, but it is the slowest and can leave areas untested if the tester never discovers them. White box is the fastest and most thorough, because the tester is not spending time on reconnaissance and can examine everything, but it is the least like a real outsider. Grey box falls in the middle and is often the most cost effective, because giving the tester a normal user’s access reflects a realistic scenario, an attacker who has already phished one account, while still covering a lot of ground. The right choice depends on what an organisation is trying to learn: realism, completeness, or a balance.

External vs internal, overt vs covert

Two further distinctions describe the tester’s position and the organisation’s awareness. External penetration testing is conducted from outside the security perimeter, taking the view as seen from the internet, and it concentrates on internet facing systems and access methods such as portals and remote access. Internal penetration testing is conducted from within the network, with the tester assuming the identity of a trusted insider, usually granted general user access, and it focuses on system configuration, authentication, access control and privilege escalation up to administrator level.

The overt versus covert distinction is about who knows. Overt testing, which NIST also calls white hat testing, is performed with the knowledge and cooperation of the organisation’s IT staff. Covert testing, also called black hat testing, is performed without the IT staff’s knowledge but with the full knowledge and permission of senior management, and its purpose is to examine the damage an adversary could cause and to test how well the defenders detect and respond, not merely to find vulnerabilities. A covert engagement is the basis of what is often called a red team exercise, which is compared with a standard test in its own right. Most organisations start with overt external and internal testing, and move to covert red teaming as their security matures.

Types by target

The third dimension is what the test examines, and this is where different standards come into play. Common targets include:

  • External and internal network testing, covering the perimeter and the internal environment.
  • Web application testing, guided by the OWASP Web Security Testing Guide, covering areas such as authentication, authorisation, session management, input validation and business logic.
  • API testing, guided by the OWASP API Security Top 10, where broken object level authorisation is the leading risk.
  • Mobile application testing, guided by the OWASP mobile application security standard and testing guide.
  • Wireless testing, examining access points and wireless access methods.
  • Social engineering, which tests the human element and user awareness through techniques such as phishing.
  • Physical testing, where a tester attempts to breach physical controls to connect to a network or access sensitive information.
  • Cloud testing, a widely used modern engagement type, though it is an industry practice rather than a category defined in the older foundational standards.

The right targets depend on where an organisation’s risk lies. A SaaS company will prioritise web application and API testing, while an office based firm might prioritise external network and social engineering. A good scoping conversation identifies the targets that matter before the test begins.

Which type of penetration test do we need?

The type needed follows from what an organisation is trying to protect and what it is trying to learn. For a business launching an internet facing application, a grey box external web application test is often the sensible default. Where the question is what an attacker who gets a foothold could do, an internal test is the priority. And where the goal is to test not just the controls but the team’s ability to detect and respond, a covert red team engagement is the right, more advanced, choice.

The practical advice is to define the objective first and let it drive the type, rather than asking for a generic penetration test and accepting whatever scope comes back. It helps to be clear about what matters most, whether that is realism, coverage, a specific application, or detection and response, and to choose the knowledge level, position and targets accordingly. A capable provider will work through this in scoping and map the engagement to a recognised methodology. Where an organisation wants help scoping the right test for its risk, Cybernion offers penetration testing.

Frequently asked questions

What are the main types of penetration testing?

Penetration tests are classified three ways. By how much the tester knows, black box (no knowledge), grey box (some knowledge) and white box (full knowledge). By the tester’s position and whether staff are aware, external versus internal, and overt versus covert. And by target, such as network, web application, API, mobile, wireless, cloud, social engineering and physical testing. Most engagements combine several of these dimensions.

What is the difference between black box, grey box and white box testing?

The difference is how much the tester knows about the target. Black box testing assumes no knowledge of the internal structure, so the tester works like an external attacker. White box testing assumes substantial knowledge, including internal detail, so it is faster and more thorough. Grey box testing falls between, with some knowledge such as user level credentials. Black box is most realistic, white box is most complete, and grey box balances the two.

What is the difference between external and internal penetration testing?

External penetration testing is conducted from outside the organisation’s perimeter, taking the view an attacker has from the internet, and it targets internet facing systems and access methods. Internal penetration testing is conducted from within the network as a trusted insider, usually with general user access, and it focuses on system configuration, authentication, access control and privilege escalation. Many organisations do both to cover the outside and inside views.

What is a covert or red team penetration test?

A covert test, which NIST also calls black hat testing, is performed without the knowledge of the organisation’s IT staff but with the full knowledge and permission of senior management. Its purpose is to examine the damage an adversary could cause and to test how well the defenders detect and respond, rather than simply to list vulnerabilities. An overt test, by contrast, is done with the IT staff’s knowledge and cooperation.

What targets can a penetration test cover?

A penetration test can target external and internal networks, web applications, APIs, mobile apps, wireless networks, and cloud environments, and it can include social engineering to test the human element and physical security to test facility controls. The right targets depend on where an organisation’s risk and its most valuable systems lie. Different targets draw on different standards, such as the OWASP guides for web, API and mobile testing.


Written by Gaurav Vikash, an ASD endorsed IRAP assessor and senior cyber security leader with 18 years of experience across Australia, the UK and Asia, including CISO and senior security leadership roles. He holds CISSP, CISA, CISM and CRISC and is an ISO 27001 Lead Auditor and ISO 42001 Lead Implementer, and speaks regularly at industry conferences.

Talk to us.

Sources:

  1. NIST SP 800-115, Technical Guide to Information Security Testing and Assessment, National Institute of Standards and Technology, September 2008
  2. Black box / gray box / white box testing (Glossary), NIST CSRC, 2026
  3. OWASP API Security Top 10 – 2023, OWASP, 2023
  4. Web Security Testing Guide, OWASP, 2026

Last updated: 30 August, 2026

Talk to us about your engagement

A scoped proposal within one business day.