Cybersecurity
What is Penetration Testing?
Also called pentest, pen testing.
Short definition
Penetration testing is security testing that simulates a real attack on a system, with the owner's written permission, to prove which weaknesses an attacker could actually exploit.
How it works
The tester starts from an agreed scope, such as one website and its admin panel. They gather information about the system, look for weak points, then try to exploit them the way an attacker would: logging in without a password, reading data that should be restricted, or escalating privileges. Each finding is backed by reproducible steps and assigned a severity level, usually based on a CVSS score.
The end result is a report containing a summary for management, technical details for each vulnerability, and remediation recommendations. After the team fixes the issues, the tester usually runs a retest to confirm the vulnerabilities are truly closed.
Example in a company
A company is about to launch a customer portal that stores personal data. Before going live, the security team requests a pentest. The tester finds that changing a number in the profile page URL opens another customer's profile. A flaw like this does not show up in ordinary automated scans, but it is fatal if an outsider finds it first.
How it differs from vulnerability assessment
A vulnerability assessment maps as many known vulnerabilities as possible, mostly through scanning. Penetration testing is deeper and narrower: it proves whether a vulnerability can be exploited and how far the impact goes. Many companies run assessments regularly, then a pentest before a major launch or once a year.
Related terms
Further reading
Resources
- NIST Glossary: penetration testing (csrc.nist.gov)
- OWASP Web Security Testing Guide (owasp.org)