A TPRM lead sends a hosting provider a finding: an exposed admin portal running a version with a known, exploited flaw. Two days later, the vendor replies that a fix is on its roadmap for Q2. Whether that answer is acceptable depends on a distinction most teams use loosely, the one between remediation and mitigation, and on who is doing what in the months before Q2 arrives. So before deciding, settle a basic question: what is remediation, and what only looks like it.
Most third-party programs lose ground in the gap between finding and fixing. Panorays’ 2025 CISO survey found that half of organizations leave 26% to 50% of identified third-party vulnerabilities unresolved, and 98% leave at least 10% unaddressed. This guide defines remediation and mitigation, shows how to choose between them when the fix sits with a vendor, and ends with a remediation plan you can send a supplier.
What is remediation?
Remediation is the work of removing a security weakness at its source so the risk no longer exists in that form. In cybersecurity, that usually means patching or upgrading software, closing an exposed service, fixing a misconfiguration, or replacing a control that failed. In third-party risk, the source is often in the vendor’s environment, so remediation is something you request, track, and verify rather than something your own team does.
A remediated finding is closed with evidence: a new version number, a scan that no longer sees the service, a revised policy, or a test that passes. Until that evidence exists, the finding is open, however confident the vendor sounds.
What is the difference between remediation and mitigation?
Remediation removes the weakness. Mitigation reduces the chance it is exploited, or the damage if it is, while the weakness is still there. Patching a flaw is remediation. Blocking the vulnerable port, limiting what data the vendor can reach, or adding monitoring until the patch lands is mitigation. Most real cases use both, mitigation first and remediation as the finish line.
| Remediation | Mitigation | Acceptance | |
|---|---|---|---|
| What changes | The weakness is removed | Likelihood or impact is reduced; the weakness remains | Nothing changes; the risk is recorded and owned |
| Who acts in TPRM | Usually the vendor | The vendor, your team, or both | Your risk owner signs off |
| Typical examples | Patch, upgrade, close the service, fix the configuration | Firewall rule, reduced data sharing, extra logging, MFA on the connection | Low-impact finding on a vendor with no access to sensitive data |
| How it closes | Evidence the weakness is gone | A review date, because the risk is still live | An expiry date and a named approver |
The difference matters most in reporting. A finding marked “mitigated” is still an open risk with a temporary control around it, and it needs a date when someone checks whether the real fix arrived. Teams that record mitigations as closed lose track of exactly the exposures that later turn into incidents.
When should you remediate, and when should you mitigate?
Remediate whenever a fix exists, and the vendor can apply it in a time that matches the risk. Mitigate when the fix does not exist yet, when it will take longer than the exposure allows, or when the vendor cannot or will not apply it. Four questions settle most cases: how severe is the flaw, is it being exploited, does a fix exist, and what can this vendor reach in your business?
Exploitation changes the clock more than anything else. Verizon’s 2025 Data Breach Investigations Report found that only about 54% of vulnerabilities in edge devices such as VPNs and firewalls were fully remediated during the year, with a median of 32 days to do it. Attackers often move much faster, and CISA’s Known Exploited Vulnerabilities catalog is the quickest public check on whether a flaw is already in use. So for an internet-facing system at a critical vendor, a month of waiting is exposure, and mitigation has to cover the gap.
What the vendor can reach is the other half. The same unpatched server means one thing at a vendor that hosts your customer data and something much smaller at a vendor that prints your event badges. Severity scores describe the flaw; the relationship describes what it costs you. Our piece on contextual risk scoring covers how to weigh the two together.
How does remediation work when the fix sits with a vendor?
Third-party remediation runs in six steps: share the evidence, set severity and a due date by vendor tier, agree a plan with the vendor, put interim mitigation in place, verify the fix, and close or formally accept what is left. The steps are the same for every vendor; the depth and the deadlines change with the tier and with whether the flaw is being exploited.
- Share the evidence. Send the vendor what you observed, when, and why it matters, in terms its engineers can act on. “Your score dropped” gets a polite reply. “Host X exposes service Y on version Z, affected by CVE-N, seen on this date” gets a ticket.
- Set severity and a due date by tier. A critical vendor with an exploited flaw might get days; a low-tier vendor with a theoretical one might get a quarter. Write the timelines down once so every finding does not become a negotiation.
- Agree a plan. Ask the vendor to confirm the fix, the owner on its side, and the date, in writing. The template below gives the fields.
- Mitigate in the meantime. Decide what you and the vendor will do until the fix lands, such as restricting the connection, reducing data sharing, or watching the integration more closely.
- Verify. Check the fix yourself where you can, through an external scan, a version check, or evidence the vendor provides. A ticket marked “done” is a claim, not proof.
- Close or accept. Close the finding with its evidence, or record a formal acceptance with an owner and an expiry date if the vendor cannot fix it.
What goes in a remediation plan?
A remediation plan is a short written agreement between you and the vendor about one finding or a group of related findings: what is wrong, what will be done, by whom, by when, and how both sides will know it is fixed. Keep it to a page. A long plan is a sign that several findings with different owners have been bundled together and should be split.
Here is a template you can send a supplier as is.
| Field | What to write |
|---|---|
| Finding | One sentence: what was observed, where, and when |
| Evidence | The scan result, questionnaire answer, or document that shows it |
| Affected service | Which service you buy from the vendor is exposed, and what data or access it involves |
| Severity and tier | Severity of the flaw, the vendor’s tier, and whether it is known to be exploited |
| Required fix | The remediation you expect, such as a patch level, a closed port, or a revised control |
| Due date | The date agreed with the vendor, based on your timeline for this severity and tier |
| Interim mitigation | What the vendor will do until the fix lands, and what your team will do |
| Owners | A named person at the vendor and a named person on your side |
| Verification | How the fix will be confirmed, such as an external rescan, a version screenshot, or a test result |
| Escalation | Who is contacted if the date slips, and the contract clause that applies |
| Exception | If the fix is not possible: the accepted risk, the approver, and the date the acceptance expires |
The interim mitigation and exception rows carry the most weight. They turn “on our roadmap for Q2” into a decision someone owns. Our article on why suppliers need a remediation plan covers the case to make to a vendor that pushes back.
How do you get a vendor to remediate?
Vendors fix what is specific, small, and tied to something they agreed to. A finding with clear evidence, a single owner, and a date in the contract gets fixed far more often than a score change or a 40-item report. The same survey found 25% of CISOs struggle with vendor relationship management, so the way you ask matters as much as what you ask for.
Four habits help. Send one finding per request, or one group with one owner, so it lands in a single ticket. Lead with the business reason and the evidence, not a rating. Write remediation timelines and notification duties into contracts at onboarding and renewal, when you have the most influence, so a request later points to a clause instead of goodwill. And give the vendor a way to dispute a finding: a false positive that stays on the record teaches the vendor to ignore the next real one.
Escalate on a schedule rather than a mood. If a critical finding passes its due date, the relationship owner hears first, then the vendor’s security lead, then procurement. Each step should already be written in the plan.
How does remediation connect to residual risk?
Remediation and mitigation are the two tools that move a vendor from its inherent risk down to its residual risk, the exposure that remains after controls. Every closed finding should lower the vendor’s residual risk; every accepted exception keeps part of it in place on purpose. Tracking findings this way lets you show leadership what was fixed, what was reduced, and what was knowingly accepted.
Third-party cyber risk management earns its keep at that point. Contextual risk scoring weights each finding by what the vendor can reach in your business, so the exposed admin portal at the hosting provider rises to the top and the same flaw at the badge printer waits its turn. Continuous monitoring then confirms whether the fix held. For the scoring side, our guide to residual risk works through an example, and our vendor risk assessment guide covers how findings are found in the first place.
Want findings tracked to closure with your vendors? See how Panorays approaches third-party cyber risk management.
Remediation vs Mitigation FAQs
What does remediation mean in cybersecurity?
In cybersecurity, remediation means fixing a vulnerability or control failure at its source so the risk is removed. Common examples are applying a patch, upgrading unsupported software, closing an exposed service, and correcting a misconfiguration. Remediation is complete only when there is evidence the weakness is gone, such as a rescan, a version check, or a passed test, rather than a statement that the work was done.
What is a remediation plan?
A remediation plan is a written agreement on how a specific finding will be fixed. It names the finding and its evidence, the required fix, the owners on both sides, the due date, the interim mitigation, how the fix will be verified, and what happens if the date slips. In third-party risk, it turns a vendor’s general promise into a dated commitment that both sides can track.
Is patching remediation or mitigation?
Applying a patch that fixes the flaw is remediation, because the weakness is removed. Steps taken while waiting for the patch, such as blocking access to the vulnerable service, turning off a feature, or adding monitoring, are mitigation. Some vendor advisories call a workaround a “mitigation” for this reason: it lowers the risk without removing the cause, so the finding stays open until the patch is applied.
What is mean time to remediate?
Mean time to remediate (MTTR) is the average time between finding a vulnerability and confirming it is fixed. For third parties, measure it per vendor tier and per severity, and start the clock when the finding is shared with the vendor. A rising MTTR at a critical vendor is an early warning that its security program, or its attention to yours, is slipping.
What happens if a vendor cannot remediate a finding?
Record it as a formal exception. Agree on compensating controls, such as reduced data sharing, a restricted connection, or closer monitoring; have a named risk owner accept the remaining risk, and set an expiry date for that acceptance. If the risk is too high to accept and cannot be mitigated, change the scope of the service or replace the vendor.