Aipedia: IT consultants and developers
When an AI-generated script or automated fix takes a client's systems down, the managed service provider usually faces an SLA claim and possibly a negligence claim. MSPs also face proposed new statutory security duties under the Cyber Security and Resilience Bill.
Part of: AI in IT consultants and developers
In short
Managed service providers use AI to triage tickets, write administration scripts, summarise alerts and trigger remediation. Because MSPs hold privileged access across many clients, one AI error can cause outages at scale. Claims usually arise under service level agreements and in negligence. The Cyber Security and Resilience Bill, now before the House of Lords as HL Bill 32, would bring medium and large MSPs into scope with duties including robust security and 24-hour and 72-hour incident reporting. Technology PI usually responds to negligent service failures, subject to the wording, with cyber cover relevant where an attack is involved.
AI is now common across the service desk. It classifies and routes tickets, drafts replies, suggests fixes from knowledge bases, summarises monitoring alerts, writes scripts for engineers to run and, in some setups, triggers automated remediation. The appeal is speed and consistency across many clients. The risk is that a confident but wrong suggestion, run with administrator rights, can affect every endpoint or tenant it touches.
There is no statutory regulator for MSPs today. BCS members are bound by the BCS Code of Conduct, including 3.1 (due care and diligence) and 3.3 (responsibility for your own work and that of colleagues you supervise). The AI Cyber Security Code of Practice is relevant where AI tools run in your environment: principle 4 (human responsibility), principle 6 (secure infrastructure) and principle 12 (monitor system behaviour).
The Cyber Security and Resilience Bill was introduced in November 2025 and is now before the House of Lords as HL Bill 32. If passed, it would bring medium and large managed service providers into scope. Duties would include robust security, telling customers about incidents, and a 24-hour initial and 72-hour full incident report. The ICO, to become the Information Commission, is named as regulator. Thresholds and registration would be set in secondary legislation, so which MSPs are covered is not yet known. See the Bill and MSPs.
Illustrative scenario: an engineer at an MSP asks an AI assistant for a script to clear old user profiles across a client's estate. The script's filter is broader than intended. It is pushed through the remote management tool without testing on a pilot group and deletes active profiles on hundreds of machines, including locally stored files. The client loses several working days, misses contractual deadlines of its own and claims service credits plus wider losses, alleging negligence.
Other patterns include AI triage downgrading a genuine security alert, and automated remediation closing a ticket without fixing the underlying fault.
An SLA often sets service credits and a liability cap. A client may argue that a reckless AI-driven change falls outside the cap or that its wider losses are recoverable in negligence. Technology PI commonly responds to claims that your managed services were negligently performed, subject to the wording, though liability you accept by contract beyond what the law would impose may be excluded. If the outage was caused or worsened by an attack, cyber cover becomes relevant. See MSP PI insurance and project and SLA claims examples.
Clients increasingly ask MSPs how AI is used in their service, and public sector buyers may ask suppliers to declare AI used in service delivery under PPN 017. A short, accurate statement of where AI assists engineers, where it acts automatically and what human checks apply helps set expectations. It also reduces the risk of a later argument that the client never agreed to automated changes in its environment, and it supports BCS clause 3.5 on not misrepresenting performance information.
Apex arranges PI and cyber cover for managed service providers. A broker reviews your SLAs alongside your cover and how your policy treats AI-driven service failures. See also deploying AI agents, AI-assisted data migration and the IT consultants AI hub.
It would bring medium and large managed service providers into scope if passed. The Bill is before the House of Lords as HL Bill 32, and thresholds and registration would be set in secondary legislation, so the exact scope is not yet confirmed.
In-scope MSPs would need to report incidents with a 24-hour initial report and a 72-hour full report, and tell customers about incidents. The ICO, to become the Information Commission, is named as regulator for in-scope MSPs.
Technology PI commonly responds to claims that managed services were negligently performed, subject to the wording, limits and any exclusions for contractual liability. If an attack was involved, cyber cover may also respond. Review both policies against your SLAs.
They may, depending on the contract. Many SLAs make service credits the sole remedy for missed service levels but leave other claims subject to a separate cap. Clients may argue that negligence falls outside the credits regime, so check the drafting.
Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.
Start a PI quote Request a callbackApex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.