08 Nov 2026 By CISO Consulting team 40 views 5 min read

CST Cybersecurity Framework: 5 Detection Coverage Mistakes

Telecom, cloud and IT providers regulated by CST often define detection coverage by tools owned rather than threats seen. This article covers five recurring mistakes, what to do instead, and how to decide which gaps to close internally and which to hand to an MDR provider.

A security analyst reviewing detection dashboards on several monitors in a dimly lit operations room.
In this article
  1. Mistake 1: Monitoring endpoints while cloud and identity logs sit unwatched
  2. Mistake 2: Treating alert volume as coverage
  3. Mistake 3: Leaving containment authority undefined
  4. Mistake 4: Skipping out-of-hours cover
  5. Mistake 5: Never testing detections against real techniques
  6. Score your own coverage
  7. Which gaps to fix internally and which to hand over
  8. How we help

Detection coverage is easy to claim and hard to demonstrate. For providers working under the CST cybersecurity regulatory framework, detection and response is more than a tooling question: it asks whether your organization can see a threat, decide what to do and act in time. This article sets out five mistakes we see repeatedly when telecom, cloud and IT service providers define their coverage, with a short "do this instead" for each, and a simple way to score yourself at the end.

The Communications, Space and Technology Commission (CST) sets the Cybersecurity Regulatory Framework (CST CRF) for the providers it regulates. We do not cite specific control numbers here. The principle running through the framework is that monitoring, incident detection and response should be effective and demonstrable, not merely present.

Mistake 1: Monitoring endpoints while cloud and identity logs sit unwatched

Endpoint agents are the most familiar control, so many programmes start and stop there. Yet for a cloud or IT service provider, a meaningful share of attacker activity happens elsewhere: a compromised administrator account, a new access key, a changed role assignment, a storage setting opened to the internet. None of these touch an endpoint.

Do this instead: list every log source that records who did what in your environment: identity provider, cloud control plane, remote access, email, network boundary and key business applications. Mark each one as collected, monitored with tested detections, or neither. Collected but unwatched counts as a gap.

Our Managed Detection and Response (MDR) service has analysts who detect threats across endpoints, network and cloud. Where the gap is configuration rather than activity, our Managed Cloud Security Posture service reviews cloud accounts and workloads continuously against your policy and the Saudi cloud controls.

Mistake 2: Treating alert volume as coverage

A busy console feels reassuring. But thousands of alerts a day do not show that the right threats are visible; they often show that detections are untuned and that analysts are triaging noise. Worse, high volume trains teams to close alerts quickly, which is where genuine intrusions get lost.

Do this instead: measure coverage by the behaviours you can detect, not by the alerts you receive. For each threat scenario relevant to your services, such as credential theft, abuse of privileged access or data staging, record which detection would fire, which data it depends on, and who investigates. Retire or tune detections that no one acts on.

An analyst-led service closes this by investigating rather than forwarding. Our MDR service offers threat investigation, not just alert forwarding, and detection tuned to your environment and threats. Each month it is reviewed for lessons learned and tuning, and we keep a detection tuning log so changes are traceable.

Mistake 3: Leaving containment authority undefined

Many providers can detect an incident and still lose hours, because nobody has decided who may disable an account, isolate a host or block an address. The analyst escalates, the on-call manager cannot reach a decision-maker, and the attacker keeps moving. In a regulated service environment, hesitation also carries operational and reporting consequences.

Do this instead: write a response authorization matrix before you need it. For each class of action, state whether it is pre-approved, needs approval from a named role, or is never taken without executive sign-off. Include contacts and a fallback for each approver.

This is the first step of how we deliver MDR: we agree response limits, contacts and approval rules at the outset, and the result is a response authorization matrix. Our service takes containment actions within limits you authorize. We act without asking only on the actions you pre-authorize in the matrix; everything else is recommended and waits for your approval.

Mistake 4: Skipping out-of-hours cover

Attackers do not follow business hours, and intrusions often begin on weekends or overnight. A rota where one engineer checks email on their phone is not detection cover; it is a hope that nothing happens while the team is asleep. For telecom and cloud providers whose customers depend on continuous service, the gap is both a security and a regulatory exposure.

Do this instead: decide the response time you need for each severity at any hour, then test whether your current arrangement meets it. If it relies on individuals being awake and reachable, count it as partial coverage at best. Building a genuine round-the-clock rota internally requires enough trained analysts to cover leave, illness and turnover, which is a real cost.

Where that is not realistic, our Managed SOC provides continuous monitoring, alert triage, escalation to your team and monthly reporting, with eyes on your environment around the clock. During an incident under MDR, you have a direct line to an analyst.

Mistake 5: Never testing detections against real techniques

A detection rule that has never fired in a test is an assumption. Log sources change, agents fail silently, parsing breaks after an update, and a rule written for one technique may miss a common variation. Programmes discover this only during a real incident or an audit question they cannot answer.

Do this instead: choose a small set of attacker techniques relevant to your services and run them in a controlled way on a schedule. For each, record whether the detection fired, how long it took, whether the analyst understood it, and what was changed afterwards. Keep the record; it is evidence of effective monitoring.

A monthly review cycle supports this discipline. Our MDR service delivers a monthly detection and response report and a detection tuning log, and each confirmed incident receives a case report with root cause and next steps.

Score your own coverage

Rate each of the five areas from 0 to 2: 0 means absent, 1 means partial or dependent on individuals, 2 means defined, tested and evidenced.

  1. Log sources: endpoint, cloud, identity and network all monitored.
  2. Detection quality: coverage described by behaviour, with tuned rules.
  3. Containment authority: a written matrix with named approvers.
  4. Out-of-hours cover: tested response at any hour.
  5. Testing: detections exercised against real techniques on a schedule.

Any area scoring 0 needs attention now. Areas scoring 1 are candidates for deliberate decisions.

Which gaps to fix internally and which to hand over

Some gaps are best closed by your own team, because they depend on knowledge of your business and decisions only you can make. Defining the authorization matrix content, deciding risk appetite, and fixing log collection at source sit mainly with you.

Others are about sustained capacity and analyst depth: round-the-clock investigation, continuous tuning and case-by-case response. These are strong candidates for an MDR provider, because the constraint is people and continuity rather than policy.

How we help

Our Managed Detection and Response service is ongoing, on 12-month terms. We agree response limits, contacts and approval rules, deploy sensors, integrations and baselines, respond to each case by investigating, containing and reporting, and review lessons learned and tuning every month. The service maps to the CST CRF, alongside SAMA CSF and NCA ECC. We confirm supported platforms during scoping, because the service is tool-aware, not tool-bound.

If you have scored your coverage and found gaps you would rather not carry alone, we can start by reviewing your results with you and agreeing what an authorized response model would look like.

Tags: CST CRF Detection and Response MDR Telecom Security Cloud Security SOC
Readiness self-check

How ready are you? Find out in two minutes

Answer a few questions for your framework. You get a score, your biggest gaps and — if you want it — a detailed assessment from our team.

Ready to talk about your compliance?

Tell us where you stand. We will show you the shortest path to what your regulator expects.

Regulatory updates in your inbox

SAMA, NCA and SDAIA changes and what they mean for your institution — once a month.

E-mails in

We confirm by e-mail; unsubscribe any time. Manage your subscription · Past issues

Schedule a Free Assessment