Part 2 of an “Uneventful by design” blog series
It's 4:47 p.m. on a Friday when the Cybersecurity and Infrastructure Security Agency (CISA) adds a new entry to the Known Exploited Vulnerabilities (KEV) catalog. The affected component is common middleware, and it runs somewhere in your environment.
Under CISA’s Binding Operational Directive (BOD) 22-01, the remediation clock is now running. The question isn't whether to patch. It's how to patch quickly while every service that depends on that component keeps meeting its service levels.
Many teams treat that as a trade-off: Move fast and accept risk to stability or move carefully and risk missing the deadline. In mission-critical operations, it doesn't have to be either. Speed and stability come from the same discipline.
Incomplete inventory: The weekend that goes wrong
In the eventful version, the team spends Friday evening working out where the vulnerable component actually runs. The inventory is partial, so they find most instances and miss some. The patch goes out broadly on Saturday to beat the clock. One application doesn't tolerate the update. By Sunday, a service that Veterans and Department of Veterans Affairs staff rely on is degraded, and the team is choosing between rolling back and leaving the vulnerability open.
The patch itself becomes the incident.
Continuous inventory: The weekend that goes right
In the uneventful version, the team answers "where does this run?" in minutes, because asset and software inventory is continuous. By Friday evening they have a list of affected systems ranked by exposure and mission impact.
The patch goes first to a small canary group with active monitoring. Health checks confirm the services behave normally. The rollout widens in stages, with a tested rollback ready at every step. By the due date, every instance is patched, verified and documented. Service levels never moved.
That is what we mean by “maintain and enhance”: The environment is more secure on Monday than it was on Friday, and nobody outside the operations team noticed the change.
Why the window keeps shrinking
The pace of federal vulnerability response keeps increasing. BOD 22-01 sets firm remediation due dates for catalogued vulnerabilities. BOD 23-01 expects regular, automated asset discovery and vulnerability enumeration. OMB M-22-09 adds zero trust policy and monitoring expectations on top. Each raises the volume of change an operations team must absorb without disrupting the mission.
Proven at enterprise scale
The belief that speed and stability compete is common in commercial IT, too. CGI's work with Volkswagen Group shows why it doesn't hold.
When Volkswagen and CGI launched a new co-governed software delivery model in 2024, they did not flip a switch. They started with a proof of concept, then moved through a phased implementation designed specifically to minimize risk and keep the transition smooth. The deliberate pace was the fast path.
The reported results make the point. Volkswagen gained predictable, sprint-based delivery cycles and faster time to market. It also gained better software quality and reduced security risk. Those outcomes arrived together, from one operating model, under a joint governance board that reviews priorities and progress every quarter. Security and delivery were measured side by side, not traded against each other.
The principles at work
Rapid vulnerability remediation often gets framed as a race against the clock. In practice, the outcome is usually decided long before a new KEV appears. Organizations that consistently remediate quickly without disrupting services tend to rely on the same underlying disciplines.
Three are especially relevant here:
- Know everything you run. The first challenge in any KEV response is identifying exposure. Continuous asset and software inventory turns hours of investigation into minutes of decision-making.
- Change slowly to move fast. Staged rollouts, canary deployments and validation checkpoints can appear deliberate, but they often accelerate remediation by preventing outages, rework and emergency rollbacks.
- Rehearse failure. Recovery plans create confidence only when they have been tested. Teams move faster when they know exactly how they will respond if an update behaves unexpectedly.
These principles are not specific to a particular technology stack, security tool or operating model. They emerge repeatedly wherever both security and availability matter. The teams that consistently meet remediation deadlines aren't relying on speed alone. They already know their environment, trust their deployment process and have practiced recovery before they need it.
What "done" really means
One of the most common mistakes in vulnerability remediation is defining success too narrowly. Applying a patch is an activity. Reducing risk without disrupting the mission is the outcome.
A KEV response is complete only when four things are true:
- Every affected instance is remediated.
- The remediation is verified, not assumed.
- The work is documented and auditable.
- Dependent services continue to meet expected performance and availability levels.
That final measure is often the most important. For the Veterans and staff relying on the mission, a vulnerability is not truly resolved if the fix creates a new operational problem. The real objective is not simply to close a security finding. It is to leave the environment more secure than it was before while maintaining the experience stakeholders expect.
Three questions worth asking
These strengths and weaknesses usually remain invisible until a remediation deadline puts leaders under pressure. A Friday afternoon KEV announcement doesn't create operational problems. It reveals the ones that already exist.
Three questions can help determine whether the foundation for fast, low-disruption remediation is already in place:
- Could we identify every affected system within the first hour of a KEV announcement? This reveals whether visibility is enabling action or slowing it down. Time spent searching for exposure is time taken away from testing, validation and controlled deployment.
- How confident are we that a failed update could be reversed immediately? Confidence comes from practice, not documentation. A tested rollback capability reduces hesitation and lets teams move quickly when timelines are compressed.
- Do we measure service performance alongside remediation performance? Security and operations ultimately produce the same outcome: mission delivery. Tracking one without the other makes it possible to declare success while shifting risk elsewhere.
The goal is not perfection. The goal is understanding where friction, uncertainty or operational risk exists before the next deadline arrives.
Uneventful by design
Fast remediation and steady service levels aren't competing goals. They are the same goal, delivered by the same discipline, and the result is a mission that keeps running for every Veteran who depends on it. CGI has proven that discipline in one of the world's largest commercial enterprises and applies it to the federal missions that serve Veterans.
Next in the series: How forecasting keeps Veteran satisfaction high on VA's busiest days.