πŸ”

AI in IT: 6 MSP Engineer Roles That Won't Exist in 2028

Technology Trends 2026-07-02 πŸ• 5 min 916 words Updated 2026-07-27

The Shift Is Already Here

In 2025, one of Australia's largest MSPs quietly reduced its Level 1 helpdesk team by 40% β€” not through layoffs, but through attrition. They simply didn't backfill roles because an AI triage system was handling first-line tickets.

In 2026, that same MSP is rolling out AI-generated documentation, automated patch management, and AI-driven network monitoring that flags anomalies before they become incidents.

This is not a prediction. This is happening now.

Here are the six roles that will be transformed or eliminated in MSPs within the next 18-24 months β€” and what to do about it.

1. Level 1 Helpdesk Engineer

Why It's Disappearing

AI triage systems can now: - Classify and prioritise tickets by severity - Answer common questions from a knowledge base - Run basic diagnostic scripts automatically - Escalate only the genuinely complex issues to humans

The math is brutal. An AI triage system costs a fraction of an L1 salary and works 24/7. It doesn't call in sick. It doesn't burn out.

What's Left

L1 roles won't vanish entirely, but they will shrink dramatically. The remaining positions will require technical depth β€” not just script-following. If your L1 role consists mainly of password resets, mailbox creation, and ticket triage, your timeline is 12-18 months.

The Escape Route

Move to L2 or specialise. Learn automation β€” if you can build the AI tools instead of being replaced by them, you've flipped the dynamic.

2. Report Generator / Data Entry

Why It's Disappearing

AI can generate client reports, compliance documentation, and monthly service reviews in seconds. Models like GPT-4 and Claude can ingest raw monitoring data and produce polished PDF reports with zero human intervention.

What's Left

Strategic analysis β€” interpreting what the data means, not just formatting it. Engineers who can tell a client "your security posture improved because of X, but you're vulnerable to Y" will be valued. Engineers who just paste numbers into a template won't.

3. Basic Network Monitor

Why It's Disappearing

AI-driven monitoring platforms (Cisco AI, LogicMonitor, Datadog) now detect anomalies, predict failures, and even auto-remediate common issues. The human-in-the-loop is increasingly optional for routine monitoring.

What's Left

Complex incident response and architecture design. The human value shifts from "watching dashboards" to "designing the monitoring system" and "handling the 5% of incidents the AI can't solve."

4. Manual Scripting / Automation Engineer (Basic)

Why It's Disappearing

Ironically, the AI that's replacing other roles is also replacing the people who write basic automation scripts. Copilot, Claude, and ChatGPT can generate PowerShell, Python, and Bash scripts from natural language prompts.

An engineer who previously spent 2 hours writing a migration script can now do it in 10 minutes with AI assistance.

What's Left

Automation architecture β€” designing the system, not writing the code. The premium shifts from "can you write a PowerShell script?" to "can you design an automated workflow that spans RMM, PSA, and client environments?"

5. Password Reset / Account Admin

Why It's Disappearing

Self-service password reset (SSPR), automated account provisioning via Entra ID, and AI-driven identity governance have made this role largely obsolete. Modern MSPs set it up once and it runs itself.

What's Left

Identity architecture and security policy design. The people who design the Entra ID tenant, configure conditional access policies, and manage privileged identity management will remain. The people who manually create accounts won't.

6. Basic Documentation Writer

Why It's Disappearing

AI can generate documentation from ticket histories, change logs, and system configurations. Tools like Notion AI, Guru, and custom LLM pipelines now produce client-facing documentation that's more consistent and up-to-date than manually written docs.

What's Left

Knowledge management architecture β€” designing the systems and workflows that feed the AI. The skill shifts from "writing docs" to "designing how knowledge flows through the organisation."

What Engineers Must Do Now

The MSP industry is not dying, but the role mix is shifting dramatically. Here's your survival plan:

Short-Term (0-6 months)

  • Automate your own job. If you can describe a task in a prompt, it can be automated. Do it before your employer does.
  • Learn AI tooling. Copilot, ChatGPT, Claude β€” become proficient. The engineers who use AI effectively will replace those who don't.
  • Document everything you do. Not for job security β€” for the portfolio that gets you your next role.

Medium-Term (6-18 months)

  • Specialise in security. This is the highest-growth, most AI-resistant area in IT. Essential 8, SOC, threat hunting, DFIR.
  • Move up the stack. From break-fix to architecture. From following runbooks to designing them.
  • Develop client skills. Relationship management, consulting, pre-sales β€” these are hard to automate.

Long-Term (18+ months)

  • Become the person who designs the AI systems, not the person they replace.
  • Build a personal brand. Write, speak, share. The engineers with visible expertise will always have options.
  • Stay portable. Don't let golden handcuffs (training bonds, non-competes) trap you in a dying role.

The Verdict

AI is going to reshape the MSP workforce. Some roles will shrink. Some will disappear. But the demand for skilled engineers who understand both technology and business is only going up.

The question isn't "will AI take my job?" β€” the question is "am I positioning myself for the roles AI creates, or clinging to the roles it eliminates?"

Frequently Asked Questions

Is AI really going to replace MSP engineers?
AI will not replace all MSP engineers, but it will dramatically reduce headcount in certain roles. L1 support, basic scripting, manual reporting, and standard provisioning are already being automated. Engineers who adapt by learning AI operations, security, and architecture will be in higher demand than ever.
Which MSP roles are most at risk from AI?
Level 1 helpdesk, manual report generation, basic PowerShell scripting, standard provisioning tasks, password resets and account management, and basic network monitoring are the most vulnerable. In some large MSPs, L1 headcount has already been reduced by 30-40% through AI triage systems.
What skills should MSP engineers learn to stay relevant?
AI ops and prompt engineering, cyber security (Essential 8, SOC, threat hunting), cloud architecture (AWS/Azure at Associate+ level), automation architecture (designing systems, not writing individual scripts), client consulting and pre-sales, and data analysis. The premium is on skills that AI cannot easily replicate.
How can MSP engineers prepare for the AI transition?
1) Automate your own job first β€” if your role can be scripted, it will be. 2) Specialise in security β€” it's the highest-growth, AI-resistant area. 3) Develop client-facing skills β€” relationship management is hard to automate. 4) Learn AI tooling β€” Copilot, ChatGPT, and agentic frameworks. 5) Build a portfolio of automated solutions.
Should I leave the MSP industry because of AI?
Not necessarily. The MSP industry is not dying β€” it's transforming. Engineers who evolve with it will find more interesting, higher-paid work. Those who resist change will be left behind. If your MSP doesn't invest in AI tooling and training for engineers, that's a red flag that they'll cut headcount instead.

Related Reading