Eight security checks to ask a provider to show
Ask an IT provider for relevant evidence of access controls, updates, reviews, recovery, alert handling, training, endpoint coverage and documentation.

A provider’s list of security products tells you what it uses. To see how it uses them, ask for evidence covering the systems and information it would look after.
The Cyber Centre’s managed-services guidance provides a basis for that discussion. These eight requests turn broad assurances into something you can check.
1. Show how administrative access is granted
Ask how its technicians get into client systems: whether they use separate admin accounts, how they sign in, who approves access and how it’s removed. Include subcontractors and emergency access.
Ask for a redacted example of an access review. Expect phishing-resistant multi-factor authentication, such as hardware security keys, on accounts with admin rights.
2. Show what happens when patching fails
Ask how the provider identifies an available update, evaluates urgency and verifies installation. Include a device that missed the update and a system that needs a restart.
Check how unsupported equipment or an application dependency is reported to you. Automatic updates can fail without anyone noticing, so ask who follows up.
3. Show an access-review result
Ask how the provider checks that user and supplier access still matches people’s current jobs. Include departures and privileged service accounts.
Ask how a finding becomes an assigned action and how closure is verified. An access review, a vulnerability scan and a penetration test answer different questions, so check which one you’re being shown.
4. Show a representative recovery test
Ask what was restored, which dependencies were needed and whether the application could use the recovered information. Check what the test excluded.
Review whether production administration can affect backup copies. An off-site copy can still be deleted with a stolen admin account. Ask whether at least one copy is offline or immutable and protected by separate credentials.
5. Show an alert’s route to a decision
Pick a typical alert, such as a sign-in from an unusual country, and ask who receives it, during which hours and how it’s escalated. Include the handoff to your business contact.
Confirm what the provider may do without checking with you first. Ask for an example of an alert that was followed through to a decision.
6. Show how training improves a work procedure
Ask for a role-specific exercise and the reporting route it tests. Check what happens when someone reports a mistake.
Look for improvements to instructions and support. Good results on a phishing exercise won’t mean staff spot every future message.
7. Show endpoint and mail coverage separately
Ask how the provider identifies devices that aren’t enrolled, are unsupported or have failed protection checks. Include servers and permitted mobile devices where relevant.
For email, ask about three things separately: SPF, DKIM and DMARC on your domain, filtering of incoming mail, and protection for the mailboxes themselves. A message can pass SPF and DKIM and still be phishing, sent from a domain the attacker owns.
8. Show the documentation you’ll receive
Request a sample service record with sensitive client information removed. It should show clearly who’s responsible for what, the important settings, any open exceptions and emergency contacts.
Check who owns and updates the records and how they’re handed back when the arrangement ends. Weigh what the provider could show, and what it couldn’t, before you sign.


