CSA CCM TVM-09
Vulnerability Prioritization

Not all vulnerabilities are created equal. Some pose a much greater risk to your cloud environment than others. To make the most of limited resources, it's critical to prioritize which vulnerabilities to tackle first based on the relative risk they pose to your specific environment and data.

Where did this come from?

This control comes from the CSA Cloud Controls Matrix v4.0.10 - 2023-09-26. You can download the full matrix here. The matrix provides a detailed mapping of security concepts and principles to various leading standards and control frameworks. It's a great resource for ensuring comprehensive cloud security coverage.

Who should care?

  • Security engineers responsible for identifying and remediating vulnerabilities
  • DevOps teams that need to balance security with speed of delivery
  • Risk managers who oversee the organization's overall risk posture
  • Compliance officers validating adherence to security standards

What is the risk?

Failing to properly prioritize vulnerability remediation can lead to:

  • High severity vulnerabilities being exploited by attackers
  • Data breaches resulting in reputational damage and regulatory fines
  • Wasted effort remediating low risk issues while critical flaws remain
  • Strained relationships between security and DevOps teams

This control helps manage those risks by ensuring remediation efforts are targeted where they'll have the biggest positive impact. However, it's not a silver bullet. Even well prioritized vulns still need to actually get fixed.

What's the care factor?

For most organizations, this should be a high priority control. Virtually every non-trivial environment will have more vulnerabilities than can be fixed immediately. Ruthless prioritization is key to keeping the risk at an acceptable level while not boiling the ocean. The exception may be very small, low-complexity environments where comprehensive remediation is feasible.

When is it relevant?

Vulnerability prioritization makes sense when:

  • You have a large and complex environment with many vulnerabilities
  • Remediation requires coordination between multiple teams
  • You're trying to establish a vulnerability management program
  • You need to demonstrate risk reduction to management

It's less critical when:

  • You have a small, well-controlled environment
  • All vulnerabilities are remediated quickly as part of the SDLC
  • There are few or no constraints on remediation resources

What are the trade offs?

Proper prioritization takes time and effort. You need to gather data on the vulnerabilities, the affected assets, and the threat landscape. This data then needs to be analyzed to determine the risk rankings. More sophisticated prioritization models require more data and analysis.

There's also a risk of getting the prioritization wrong, either by underestimating the risk of a vulnerability or by failing to account for attacker behavior and motivations. Overly simplistic models like CVSS don't account for whether a vulnerability is being actively exploited in the wild.

How to make it happen?

  1. Integrate vulnerability scanning across development and production
  2. Aggregate vulnerability data into a single system of record
  3. Enrich the data with asset criticality and threat intelligence
  4. Use a model like DREAD to assess risk based on multiple factors:
    • Damage potential
    • Reproducibility
    • Exploitability
    • Affected users
    • Discoverability
  5. Generate a prioritized list of vulnerabilities ordered by risk score
  6. Feed the prioritized list into the remediation workflow tools
  7. Monitor and track remediation progress over time
  8. Adjust model weights and inputs based on real-world results

What are some gotchas?

  • The vulnerability scanners need read access to all systems and configs
  • Asset inventory and criticality data needs to be complete and up-to-date
  • Threat intelligence feeds may require additional licensing costs
  • The prioritization system needs to integrate with the ticketing systems
  • Remediation SLAs and policies need to align with risk rankings

What are the alternatives?

  • Prioritize solely based on CVSS scores (AWS Inspector provides these)
  • Prioritize based on asset criticality or data classification
  • Prioritize based on active exploitation in the wild
  • Adopt a zero-trust approach and assume all vulns will be exploited
  • Accept the risk and rely on defense-in-depth to limit blast radius

Explore further

Blog

Learn cloud security with our research blog