Definition
A vulnerability assessment finds and lists known weaknesses in your systems, usually with automated scanners, and ranks them by severity. Unlike a penetration test, it does not try to exploit what it finds, so it shows what might be wrong rather than what an attacker could actually do.
What a vulnerability assessment is
A vulnerability assessment is a systematic check of your systems for known weaknesses: missing patches, outdated software, insecure settings and exposed services. It is usually run with scanners that compare what they find against databases of known vulnerabilities, then rank the results, often by CVSS score.
The result is a list. It tells you what might be wrong across a large number of systems, quickly and at low cost. What it does not tell you is which of those issues an attacker could actually use.
How a vulnerability assessment works
- Define the scope: which networks, hosts and applications are included.
- Discover assets: find the systems and services inside that scope.
- Scan: check each asset against known vulnerabilities and insecure configurations.
- Analyse: remove obvious false positives and rank the rest by severity.
- Report: list the findings with suggested fixes.
Vulnerability assessment vs penetration testing
Guidance such as NIST SP 800-115 treats these as different techniques: one identifies weaknesses, the other tries to exploit them.
| Vulnerability assessment | Penetration testing | |
|---|---|---|
| Goal | List known weaknesses | Show what an attacker can actually do |
| Method | Mostly automated scanning | Attempts real attacks, step by step |
| Exploits findings | No | Yes, within an agreed scope |
| Chains weaknesses | No | Yes, into realistic attack paths |
| False positives | Common | Rare, because findings are proven |
| Typical result | A long list ranked by severity | Confirmed findings with evidence |
A simple way to remember it: a vulnerability assessment finds the unlocked windows, a penetration test climbs through one and shows how far it gets.
Limits of a vulnerability assessment
- Version checks can be wrong. Scanners often judge by the version a system reports, and patched packages can keep an old version number.
- No proof of impact. A "critical" result may not be reachable or exploitable in your environment, while a "medium" one may be the first step of a serious attack.
- Single issues only. Real attacks combine several weaknesses. A list of separate findings does not show that.
- Logic flaws are missed. Scanners do not understand how your application is supposed to work.
When to use each
Vulnerability assessments are good for frequent, broad hygiene checks across many systems. Penetration tests answer the harder question of real risk and are often required by customers and frameworks. Most teams need both: scanning to keep the list short, and penetration testing to prove which items matter. Automated and continuous penetration testing now make the second much more frequent than a yearly engagement.
How Plainsea fits in
Plainsea goes past the list. Its AI agents discover what is in scope, attempt real attacks within the limits you set and report only findings confirmed as exploitable, with the evidence and the attack path behind each one. Learn more in AI penetration testing, or see a focused example in the free PHP2Shell check.