Cybersecurity
What Is Tested in Website Penetration Testing? OWASP Top 10 for Managers
OWASP Top 10:2025 is a list of the ten most important web application security risk categories. This guide translates each category into management language: what is tested, and why it matters to the business.
SEO and Content Specialist
4 min read Reviewed by Legia Cybersecurity Team
Short answer: website penetration testing generally uses the OWASP Top 10 as a minimum checklist. Its latest edition, OWASP Top 10:2025, begins with broken access control (users can view or change data that is not theirs), security misconfiguration, and software supply chain failures, followed by cryptographic failures, injection, insecure design, authentication failures, data integrity failures, logging and alerting failures, and the mishandling of exceptional conditions. The category that most often has a major business impact is the first one.
Penetration testing reports are full of technical terms, and management often struggles to judge which ones really matter. The OWASP Top 10 helps because it groups flaws into ten widely known categories. If you understand what each category means in business terms, a test report becomes much easier to read and prioritize.
What the OWASP Top 10 is
OWASP (Open Worldwide Application Security Project) is an open community in application security. The OWASP Top 10 is an awareness document listing the ten most important web application security risk categories, and it is updated periodically. Its latest edition is OWASP Top 10:2025. It is not a certification standard, but it is a widely used starting point for testing and secure development.
The ten categories in management language
| Code | Category | What it means for the business | Example tested |
|---|---|---|---|
| A01 | Broken Access Control | Users can view or change data that is not theirs | Changing a number in the page address to open another user’s data |
| A02 | Security Misconfiguration | Default or incorrect settings open a gap | An exposed admin page, error messages that reveal system information |
| A03 | Software Supply Chain Failures | Third-party components become an entry point | Outdated libraries or plugins with known flaws |
| A04 | Cryptographic Failures | Sensitive data is not properly protected | Data sent without encryption, passwords stored in a weak way |
| A05 | Injection | User input is executed as commands | A search form that can be used to read the contents of the database |
| A06 | Insecure Design | The flaw is in the design of the flow, not in the code | A password reset flow that can be guessed |
| A07 | Authentication Failures | The login process is easy to break | No limit on login attempts, sessions that never expire |
| A08 | Software or Data Integrity Failures | Updates or data can be changed without detection | Automatic updates without an authenticity check |
| A09 | Security Logging and Alerting Failures | Attacks are not recorded or no one is notified | Repeated failed login attempts with no alert |
| A10 | Mishandling of Exceptional Conditions | The system reacts unsafely when something unexpected happens | An error that makes the system skip a security check |
The categories that most often have a major impact
Access control (A01) ranks first, and in practice it is the finding that is most often serious. Websites and applications with user accounts, such as customer portals, submission systems, or partner dashboards, are prone to this flaw: one user can see another user’s data just by changing a request slightly. Flaws like this often slip past automated scanning because they depend on application logic, so they are most effectively tested with a grey box approach: the tester uses an account for every role. The approach is explained in the article Black Box, Grey Box, or White Box.
Misconfiguration (A02) and supply chain (A03) are often found on websites that use a CMS and plugins: old versions that are not updated, plugins no longer maintained by their creators, or default settings that were never changed. The fix is relatively simple, but only if someone checks regularly.
What the OWASP Top 10 does not cover
The Top 10 is a minimum list, not a complete one. It does not cover every type of flaw, and it does not assess anything outside the web application, such as office network security or employee habits. For more detailed checks, OWASP publishes the Application Security Verification Standard (ASVS), a list of security requirements that can be verified one by one.
How to use these categories when reading a report
- Group findings by category. If many findings fall in one category, the problem is likely in development habits, not a single line of code.
- Prioritize access control and authentication on systems that hold personal data or transactions.
- Ask about the fix pattern. An injection finding in one form almost always means other forms need to be checked.
- Ask for a retest before a finding is declared closed.
How to read each part of a report is covered in the article How to Read a Penetration Testing Report.
How Legia does it
Legia’s Penetration Testing currently focuses on websites and web applications, using the OWASP Top 10 and OWASP ASVS as working references and CVSS to rate severity. These references are not certifications, but they make every finding comparable and re-checkable by another team. Testing is run manually with written permission, generally 3 to 6 weeks, and a retest after fixes is included in the package.
Frequently asked questions
Does passing the OWASP Top 10 mean the website is secure?
Not necessarily. The Top 10 is a list of the most important risks, not a guarantee. It is a good starting point, especially for a website that has never been tested at all.
How often does the OWASP Top 10 list change?
The list is updated every few years based on data and community input. The previous edition was published in 2021, and the latest in 2025.
Is automated scanning enough to test the OWASP Top 10?
Not for every category. Scanners find many misconfigurations and outdated components, but access control and design flaws are usually found only through manual testing.


