Join our Newsletter — 33% off our NHI Course
Home› Guides› Managed Service Provider Risk Assessment Guide: Score, Contract…
Guide Third-Party & Supply Chain Risk

Managed Service Provider Risk Assessment Guide: Score, Contract and Monitor Your MSP

PDF for newsletter subscribers · free to read online

← All guides
By NHI Mgmt Group, edited by Lalit Choda Updated 27 October 2026 13 min read
On this page

A managed service provider usually holds standing administrative access to the systems it runs for you, so its weaknesses become your weaknesses the moment the contract is signed. Joint advisories from the Five Eyes cyber agencies, the UK NCSC and CISA all treat MSP access as a privileged pathway that customers must assess, contract for and keep watching. This guide is written for the customer side: IT, security, procurement and risk teams who need to decide whether an MSP is safe enough, what to put in the contract and how to monitor the relationship until exit. It complements our guide for providers themselves, which covers how MSPs should secure their own estate.

Key takeaways

  • Assess the access, not just the company. The most important question is which accounts, tools and connections the MSP will hold in your environment, and how each one is protected and logged.
  • Tier MSPs by the privilege they hold. An MSP with domain admin, RMM agents on every endpoint and backup control belongs in your highest risk tier, whatever its size or certifications.
  • Certifications are a floor. Cyber Essentials Plus, ISO/IEC 27001 and SOC 2 reports show a baseline exists; they do not show that the services delivered to you are configured securely.
  • Write the controls into the contract. MFA on all MSP accounts, no reuse of admin credentials across customers, log access and retention, incident notification times and exit obligations should be contractual, not assumed.
  • Exit is part of the assessment. The joint advisory notes that disabling MSP accounts is often overlooked when a contract ends; plan offboarding before onboarding.

Why MSP risk is different from ordinary vendor risk

Most third parties hold your data. An MSP holds your keys. It typically operates with privileged identities inside your directory, remote monitoring and management (RMM) agents that can run code on endpoints, delegated administration in cloud tenants and sometimes control of the backups you would rely on to recover. A single compromise of the provider can therefore reach many customers at once, which is why the joint MSP advisory from NCSC-UK, ASD's ACSC, CCCS, CISA, NSA and FBI addresses both providers and customers.

  • Trusted relationship abuse: MITRE ATT&CK T1199 Trusted Relationship describes adversaries reaching victims through third parties with legitimate access. The APT10 Threat Actor Profile Guide covers a group known for targeting managed IT providers.
  • Remote access tooling: RMM and remote support tools map to T1219 Remote Access Tools and are a common route for ransomware distribution once a provider is breached. See the RMM Tool Security Guide.
  • Valid accounts: MSP technician and service accounts are attractive because their activity looks normal (T1078 Valid Accounts, T1133 External Remote Services).
  • Help desk and identity resets: outsourced service desks that reset credentials are a social engineering target. In the Marks and Spencer Cyberattack 2025, attackers got in by impersonating a third-party user, with reporting linking the weak point to identity verification during resets at the outsourced help desk.

For the provider-side controls, see the Cybersecurity for MSPs and MSSPs Guide.

Scoping and tiering the MSP relationship

Start by recording what the MSP will actually do and what it will be able to touch. Inherent risk follows from privilege and blast radius, not from contract value.

FactorLower riskHigher risk
Privilege heldNamed user accounts with scoped rights on a few systemsDomain or global admin, tenant-wide delegated admin, privileged service accounts
Tooling in your estateCustomer-controlled remote access with session approvalMSP-controlled RMM agent on every endpoint and server
Backups and recoveryCustomer owns backups; MSP has operator rightsMSP controls backup platform, credentials and deletion rights
Security functionsNone, or monitoring onlyMSP runs EDR, SIEM or identity provider configuration
Multi-tenancyDedicated tooling and accounts for your organisationShared RMM instance and shared technician identities across clients
SubcontractingNo subcontractors with accessOffshore or third-party technicians with access to your systems
Data exposureNo access to regulated dataAccess to personal, financial or health data

Any MSP with high-risk answers in the first three rows should be treated as a critical supplier: full assessment before contract, contractual security schedule, continuous monitoring and annual reassessment. The Vendor Tiering and Inherent Risk Questionnaire Template provides a scoring sheet, and the CIS Control 15 Service Provider Management Guide explains how classification drives assessment depth.

Assessment questions that matter

Generic questionnaires waste effort on MSPs. Focus on the domains where the joint advisory, NCSC and CISA guidance concentrate, and ask for evidence rather than yes or no answers.

DomainQuestions to askEvidence to request
Technician identityDoes every technician use a named account with MFA for all access to your environment? Are admin credentials ever reused across customers?Account list for your tenant, MFA enforcement report, tiered admin model description
PrivilegeIs access tiered, time-bound and limited to the systems the MSP manages? Are MSP accounts kept out of your internal admin groups?Role design, just-in-time access configuration, group membership export
Remote access pathIs MSP traffic restricted to a dedicated VPN or equivalent? Which RMM and remote support tools are used, and how are their consoles protected?Connectivity diagram, RMM console MFA and IP restriction settings
SegregationHow are your data and systems separated from other customers' and from the MSP's corporate network?Architecture description, tenant isolation design
LoggingWhat is logged, where, for how long, and can you access it directly? The joint advisory recommends storing the most important logs for at least six months.Log source list, retention settings, sample of your MSP access logs
PatchingWhat is the update policy for the systems they manage, and how are known exploited vulnerabilities prioritised? NCSC suggests patching critical or high-risk vulnerabilities within 14 days.Patch compliance reports for your estate
BackupsAre backups automatic, tested, isolated from the network and protected with separate keys? Who can delete them?Restore test results, backup immutability settings, list of accounts with deletion rights
Incident responseWhat happens if the MSP itself is breached? How quickly will you be told?Incident response plan extract, notification commitment, last exercise date
Personnel and subcontractorsHow are staff and subcontractors vetted? Will you be told about subcontracts that expose your data?Screening policy, subcontractor list
AssuranceWhich certifications cover the services you buy, and what is in scope?Cyber Essentials Plus certificate, ISO/IEC 27001 certificate and Statement of Applicability, SOC 2 Type 2 report

When reading assurance reports, check that the scope covers the delivery teams and tooling serving you, and read the exceptions and complementary user entity controls. See the How to Read a SOC 2 Report Guide and the Security Questionnaires Guide.

Where machine identities and secrets fit

The highest-risk MSP access is often not a person. RMM agents run as SYSTEM or root, backup platforms hold service account credentials for every protected system, and cloud integrations use application registrations, API keys or delegated admin relationships that persist long after the technician who created them has left.

  • Inventory every non-human identity the MSP uses: service accounts, app registrations, API tokens, RMM agent identities and backup credentials, each with an owner on both sides.
  • Prefer scoped, revocable mechanisms: in Microsoft 365, granular delegated admin privileges instead of broad legacy delegation. See the GDAP and MSP Delegated Admin Access Guide.
  • Ask how MSP-held secrets are stored and rotated: a provider that keeps your admin passwords in a spreadsheet or shared password vault without per-customer separation fails this test. See the Secrets Management Guide.
  • Plan revocation: you should be able to disable every MSP identity within hours if the provider reports a breach.

Contract and responsibility controls

CISA's Risk Considerations for Managed Service Provider Customers and the NCSC's guide to choosing an MSP both stress a clear responsibility matrix. Put these into a security schedule.

  • Responsibility matrix: who owns hardening, patching, detection, incident response, backup and recovery for each service.
  • Identity requirements: MFA on all MSP accounts and purchased services, named accounts, no reuse of admin credentials across customers, access limited to managed systems.
  • Logging and visibility: log retention periods, your right to direct access to security logs, and notification of confirmed or suspected events.
  • Incident notification: timeframes for notifying you of incidents affecting the MSP itself, and support for technical reports your insurer may need.
  • Service levels: separate operational and security SLAs. NCSC offers starting points for SMEs: response in about one business day for general requests and under one hour for urgent issues.
  • Change of control and subcontracting: notice of ownership or leadership changes and of any subcontract that exposes your data, with the MSP accountable for subcontractors.
  • Software transparency: CISA suggests requesting a software bill of materials or equivalent for software the MSP supplies.
  • Exit: return of data and configuration, handover of credentials, revocation of all MSP identities and confirmation of deletion of your data and secrets.

The Third-Party Security Contract Clauses Template provides wording you can adapt.

Running the assessment step by step

  1. Collect requirements before procurement. CISA recommends building a requirements master list with the CIO, CISO, operations, continuity and risk leaders, validated before vendor selection.
  2. Tier the relationship using the inherent risk factors above and decide the depth of assessment.
  3. Issue a focused questionnaire covering the ten domains, with evidence requests.
  4. Hold a technical session where the MSP walks through how a technician reaches your systems end to end, including RMM consoles and backup access.
  5. Score residual risk per domain and record gaps as contract requirements, compensating controls on your side or accepted risks with an owner.
  6. Negotiate the security schedule before signature; gaps are far harder to close afterwards.
  7. Onboard with least privilege: create MSP accounts in a dedicated container or group, outside internal admin groups, with MFA and conditional access from the start.
  8. Pilot before full access: start with a limited set of systems, confirm logging and alerting on MSP activity work, then widen scope.
  9. Reassess annually for critical MSPs and after any MSP incident, merger or major service change.

Rollback: keep a tested runbook to revoke MSP access quickly: disable accounts, remove delegated admin relationships, block RMM tooling and rotate shared secrets. The Compromised MSP Playbook covers the response in detail.

Decision guidance: small and large customers

  • Small organisations (the audience of the NCSC SME guide) often cannot audit a provider. Rely on Cyber Essentials Plus as a minimum, insist on MFA for MSP access, ask for regular health reports (patch compliance, backup success rates, unnecessary admin rights) and keep an independent copy of critical backups.
  • Mid-size organisations should add a contractual security schedule, direct log access and an annual technical review of MSP access paths.
  • Large or regulated organisations should treat critical MSPs as material third parties under their regulatory regime, require independent assurance covering the specific service, integrate MSP telemetry into their own SOC and test exit plans.
  • Choosing between MSPs: prefer the provider that gives you per-customer tooling, customer-controlled break-glass and direct log visibility over one with more certifications but shared technician identities.

Monitoring the relationship and measuring success

CISA recommends integrating MSP logs and telemetry into your own monitoring and baselining normal MSP behaviour. Alert on MSP sign-ins outside agreed hours or locations, new MSP accounts or app registrations, privilege changes, RMM tool installation on unexpected hosts, and backup deletion or retention changes.

  • Coverage: percentage of MSP identities (human and non-human) in your inventory with a named owner.
  • MFA enforcement: percentage of MSP interactive accounts with enforced phishing-resistant MFA.
  • Standing privilege: number of MSP accounts with permanent admin rights versus time-bound access.
  • Log visibility: percentage of MSP access paths whose logs reach your SIEM, and retention against contract.
  • Reassessment currency: critical MSPs assessed within the last 12 months.
  • Revocation time: measured time to disable all MSP access in the last exercise.

Framework mapping

ActivityNIST CSF 2.0NIST SP 800-53 Rev. 5CIS Controls v8.1ISO/IEC 27001:2022 Annex A
Inventory and tier MSPsGV.SC-04SR-315.1, 15.3A.5.19
Due diligence before contractGV.SC-06SR-615.5A.5.19, A.5.21
Security requirements in contractGV.SC-05SA-915.4A.5.20
MSP accounts, remote access and personnelGV.SC-07AC-2, AC-17, PS-715.4, 6.5A.5.18, A.8.2
Monitoring and log retentionGV.SC-07AU-1115.6A.5.22, A.8.15
Exit and decommissioningGV.SC-10SA-915.7A.5.22

For a structured supplier programme, see the NIST SP 800-161 Cyber Supply Chain Risk Management Guide.

Common failure patterns

  • Questionnaire without a walkthrough. The MSP answers "MFA: yes" while its RMM console or shared technician account has none.
  • MSP accounts in internal admin groups. Both the joint advisory and CISA advise against this, yet it remains the default in many onboardings.
  • Help desk resets without strong verification. Outsourced service desks that reset passwords or MFA on a phone call are a direct path to privileged access, as the Marks and Spencer attack illustrates.
  • Backups controlled solely by the MSP. If the provider is compromised, the attacker may be able to delete the copies you need to recover.
  • No log access. Without contractual rights to MSP access logs, you cannot investigate an incident involving the provider.
  • Forgotten exit. Accounts, app registrations and RMM agents outlive the contract.

Lessons from incidents

The Marks and Spencer attack in April 2025 began with the impersonation of a third-party user and, according to the reporting summarised on our breach page, cost the retailer an estimated £300 million in profit before recoveries. The lesson for MSP assessment is that the provider's identity processes, especially how its service desk verifies people before resetting passwords or MFA, are part of your attack surface. Ask the MSP to describe its reset verification step by step, require call-back or in-person verification for privileged accounts, and log every reset performed on your accounts. See the Vishing and Help Desk Social Engineering Awareness Guide.

Practitioner checklist

  • List every account, tool, connection and non-human identity the MSP holds in your environment.
  • Tier each MSP by privilege, tooling, backup control and data exposure.
  • Run a focused assessment with evidence requests and a live access walkthrough.
  • Confirm MFA on all MSP accounts, no credential reuse across customers and accounts kept out of internal admin groups.
  • Write responsibilities, log access and retention, incident notification and exit duties into the contract.
  • Restrict MSP connectivity to a dedicated, logged access path.
  • Keep an independent, tested backup that the MSP cannot delete.
  • Feed MSP activity logs into your own monitoring with specific alerts.
  • Rehearse revoking all MSP access and record the time taken.
  • Reassess critical MSPs every year and after any provider incident.

Standards and references

Related NHI Mgmt Group resources: Cybersecurity for MSPs and MSSPs Guide · Third-Party Risk Management Guide · Compromised MSP Playbook · Third-Party Access Guide

PDF for subscribers

Take this guide with you

Every section, table and checklist in one PDF, with links to the sources and related guides. Subscribe to our weekly NHI & AI Identity Journal and we’ll email you a download link.

Free weekly newsletter

Subscribe to the NHI & AI Identity Journal

The latest on NHI and Agentic AI security – articles, research, breaches, news and events every week.

Bonus 33% off our NHI Course when you subscribe.

NHIMG Editorial Note
Researched and written by the NHI Mgmt Group editorial team using AI tools, under the editorial direction of Lalit Choda. Facts are drawn from the sources listed on this page and checked by our automated quality process. Last updated 27 October 2026. Spotted an error? Use our contact page.
NHI Mgmt Group, the independent authority on Non-Human Identity, IAM and Agentic AI security. nhimg.org

PDF Downloads for Newsletter Subscribers

Get the guide

Enter your details and we’ll email you a download link.

The full guide is free to read online. The PDF is a thank you for subscribing.

Fill in all fields marked * and tick the PDF box to continue.