Findings and remediation
Every assessment finds something. What matters is the clock that starts the day a finding is discovered, the POA&M that tracks it, and the retest that closes it.
Last updated: 2026-11-22.
How findings get documented
When the 3PAO's testing turns up a weakness, it goes into the Security Assessment Report's Risk Exposure Table and then onto the Plan of Action and Milestones (POA&M). The POA&M is the working record for the rest of the authorization's life: every open finding lives there until it is closed with evidence.
Each POA&M entry carries a unique ID, the affected control, severity (aligned to CVSS), the asset involved, an owner, milestones, a scheduled completion date, a status, and the closure evidence once the fix is verified. CSPs must use the FedRAMP POA&M template, and the POA&M is updated monthly as part of continuous monitoring reporting.
The clocks
Remediation deadlines run from the day a finding is discovered, not from the day the report is written. Under the Rev5 model:
| Severity | Remediate within | Notes |
|---|---|---|
| Critical / High | 30 days | Must be remediated and retested before authorization is granted. FedRAMP will not grant certification with open High findings. |
| Moderate | 90 days | Can remain open at authorization, tracked on the POA&M. |
| Low | 180 days | Can remain open at authorization, tracked on the POA&M. |
| KEV-listed vulnerabilities | Per CISA catalog due dates | Vulnerabilities on CISA's Known Exploited Vulnerabilities list follow the catalog's own due dates rather than the severity clocks. |
The 30-day clock on High findings is the one that shapes assessment planning. A High finding discovered late in the testing window can push the authorization decision back by a month or more, because the fix has to be made, evidenced, and retested before the package goes to the Authorizing Official.
The remediation cycle
Closing a finding follows the same loop every time:
- Fix it in the federal environment. Patch, configuration change, or code deploy, in the system inside the authorization boundary.
- Produce evidence. Rescan output, configuration screenshots, attestations, whatever shows the fix is real and in place.
- Update the POA&M. Record the fix, the evidence, and the new status.
- Have the 3PAO verify closure. The assessor confirms the fix at the next assessment. A finding is not closed because the CSP says so; it is closed because the assessor verified it.
Advisory firms and GRC platforms are the main commercial help for this work. The assessor cannot do it: independence rules bar the 3PAO from remediating findings on a system it assesses.
Common findings
The same remediation work shows up across assessments. The recurring items are FIPS-validated cryptography, centralized audit logging, multi-factor authentication, configuration baselines, and vulnerability-management tooling. Most of these are cheaper to build in during readiness than to retrofit after the assessment finds them missing, which is why the readiness phase matters more than it looks.
Deviation requests
When a CSP cannot meet a requirement on schedule, it can ask FedRAMP for an approved deviation, sometimes called an operational requirement. Two limits matter. FedRAMP will not approve a deviation for a High vulnerability. And high-risk vendor dependencies must be mitigated to Moderate through compensating controls within 30 days. A deviation buys time for lower-severity items; it does not waive the hard requirements.
What happens when deadlines slip
The ConMon Playbook v1.0 (November 17, 2025) ties overdue POA&M items to graduated escalation. The published thresholds work like this: five or more high-impact findings aged past 30 days triggers one escalation level, and past 60 days a stricter one; ten or more moderate findings past 90 or 120 days escalate similarly. Missing deadlines without an approved deviation request can end in suspension or revocation of authorization. The escalation ladder is the reason the POA&M gets monthly attention rather than annual attention.
How 20x changes remediation
Under FedRAMP 20x, the flat 30/90/180-day model gives way to a likelihood × impact matrix. Fixes can be as fast as 2 days for likely-exploited high-impact flaws, and vulnerabilities must be categorized as "accepted" within 192 days at the outer edge. Findings must be evaluated within 2 days (High), 5 days (Moderate), or 7 days (Low) of detection. Detection also widens beyond scans: threat intelligence, bug bounties, supply-chain monitoring, and vulnerability disclosure all feed in.
The practical effect is that remediation under 20x is continuous rather than cyclical. Teams that treat findings as a quarterly batch will find the clocks do not wait.
Frequently asked questions
Can a service be authorized with open findings?
Yes, with limits. Moderate and Low findings can remain open at authorization, tracked on the POA&M with their clocks running. High and Critical findings must be remediated and retested first.
Who verifies that a finding is fixed?
The 3PAO, at the next assessment. Closure requires evidence of remediation verified by the assessor, not just the CSP's word.
What is a compensating control?
An alternative safeguard that reduces risk when the standard control cannot be implemented as written. FedRAMP accepts compensating controls for some situations, for example mitigating a high-risk vendor dependency to Moderate within 30 days, but not as a way around High vulnerabilities.
Remediation never really ends: once authorized, findings keep arriving through monthly scans. The continuous monitoring page covers what that looks like year after year.