# Compensating Controls: When You Can't Fix It the Normal Way

A requirement says every system must use multi-factor authentication (MFA).

Then someone points out that one older system cannot support it. It does not have the feature, and replacing it is a year away.

So what now? Ignore the requirement? Stop using the system? Often there is a third option: protect it a different way, and write that down.

---

## What a Compensating Control Is

NIST defines compensating controls as controls implemented in place of the controls in a baseline that provide equivalent or comparable protection for a system or organization.

Plain version: **when you cannot implement the expected control, you use a different one that reduces the same risk.**

A baseline is the standard set of controls expected for a system, such as the ones in NIST Special Publication (SP) 800-53B. If a baseline control is not practical, a compensating control can stand in for it.

The key words are "equivalent or comparable." A compensating control is a substitute that does the same job, not an excuse.

---

## A Simple Example

Say the requirement is to apply security updates promptly. One legacy system cannot be patched because the application running on it breaks.

Alternative protections might include:

- Network segmentation so only a few systems can reach it
- Tighter firewall rules around its ports
- Extra logging and monitoring on that system
- Restricting who can log in

None of these fix the underlying weakness. They reduce the chance or consequence of someone reaching and abusing it.

This is a good moment to remember [Patch Tuesday: How to Prioritize Vulnerabilities Instead of Patching Everything](ADD-HASHNODE-URL). Exposure and context matter as much as the patch itself.

---

## What Makes a Compensating Control Legitimate

A control becomes compensating when it is deliberate and documented. Strong entries usually answer these questions:

| Question | Why it matters |
| --- | --- |
| What requirement can't we meet? | Names the gap precisely |
| Why can't we meet it? | Shows a real constraint, not convenience |
| What are we doing instead? | Describes the alternative |
| Why is it comparable? | Explains how it addresses the same risk |
| Who owns it? | Gives someone accountability |
| When does it expire or get reviewed? | Prevents a forever workaround |
| How do we know it works? | Connects to control testing |

That last row matters. A compensating control that nobody checks is an assumption.

---

## The Temporary Access Problem Applies Here Too

In [Temporary Access Has a Way of Becoming Permanent](ADD-HASHNODE-URL), the pattern was purpose, owner, scope, expiration, and review.

Compensating controls follow the same life cycle. They are often created during a tough moment, with the intention to fix things properly later. Without an expiration and a review, "later" never arrives, and you build security debt. See [Security Debt Compounds Too](ADD-HASHNODE-URL).

A good compensating control record says what must happen for it to retire. For example: "Remove when the legacy system is replaced."

---

## Common Misconceptions

### "It is the same as ignoring the requirement."

No. Ignoring leaves the risk untreated. A compensating control treats the same risk another way.

### "Any extra control counts."

It has to address the same risk. Adding a security camera does not compensate for missing MFA.

### "It fixes the vulnerability."

Usually it does not. It reduces exposure or consequence while the underlying weakness remains. That remaining weakness may still need to be tracked in your [risk register](ADD-HASHNODE-URL).

### "Once approved, it is permanent."

Systems and threats change. Review it.

### "Compensating controls and risk acceptance are the same thing."

They are related but different. A compensating control reduces risk. Risk acceptance is a decision to live with what remains. Often you do both: add a compensating control, then accept the residual risk. See [Risk Accepted](ADD-HASHNODE-URL).

---

## Practical Exercise: Write a Compensating Control Record

This is a documentation exercise using a made-up scenario in your own lab.

Scenario: you run a small test VM with an old service you cannot update because a lab tool depends on it.

Fill in this template:

```markdown
**Requirement not met:** Keep software on current supported versions.
**Why not met:** Lab tool breaks on newer versions.
**Compensating controls:**
- Service reachable only from my admin VM
- Firewall rule limiting access to one port
- Logs reviewed weekly
**Why comparable:** Reduces who can reach the service and improves detection.
**Owner:** (you)
**Review date:** (30 to 90 days)
**Retire when:** The lab tool is replaced.
**Residual risk:** Weakness still exists; tracked as R-00X.
```

Then ask: could someone else read this and understand exactly what is protecting the system and when it ends?

---

## Key Takeaway

A compensating control is a documented alternative that provides comparable protection when the expected control cannot be used.

It is only credible when it has an owner, a reason, a way to verify it works, and a date to revisit.

---

## Verified References

- NIST Computer Security Resource Center, Glossary: Compensating Controls. <https://csrc.nist.gov/glossary/term/compensating_controls>
- NIST Computer Security Resource Center, Glossary: Compensating Security Control. <https://csrc.nist.gov/glossary/term/compensating_security_control>
- NIST SP 800-53B, Control Baselines for Information Systems and Organizations. <https://csrc.nist.gov/pubs/sp/800/53/b/upd1/final>
- NIST SP 800-53 Rev. 5 (current release 5.2.0). <https://csrc.nist.gov/publications/detail/sp/800-53/rev-5/final>

---
