Will AI Replace IT Support Workers? The 2026 Risk Picture
IT support workers score 7.5/10 on AI exposure - one of the more exposed technical roles, and a case of an industry facing disruption from within. AI-powered helpdesk chatbots, automated diagnostic tools, and self-healing IT systems are eliminating the bulk of tier-1 and tier-2 support work. BLS OEWS May 2025: 879,000 US computer support specialists at $60,810 median annual wage. ILO ILOSTAT 2024: approximately 35 million workers in ICT technician roles globally.
Key findings
- ISCO 35 (ICT and communications technicians) scores 7.5/10 on AI exposure
- Tier-1 helpdesk is largely automated: AI chatbots resolve 60-80% of routine tickets without human escalation (Gartner 2025)
- US: 879k computer support specialists at median $60,810 (BLS OEWS May 2025)
- Global: ~35M workers in ICT technician roles (ILO ILOSTAT 2024)
- Robotics risk: 2.5/10 - IT support is an AI threat, not a robotics threat
- Network/infrastructure technicians and cybersecurity ops retain stronger employment prospects
The IT sector eating its own support workers
The central irony of AI and IT support: the technology sector that builds, sells, and deploys AI tools is also the sector most rapidly automating its own support workforce. Gartner's 2025 IT support automation survey found that organisations with AI-powered helpdesk platforms resolved 60-80% of tier-1 tickets without any human involvement. ServiceNow, Zendesk, and Freshdesk all now ship AI-first support routing as standard.
The pattern is straightforward. Tier-1 helpdesk - password resets, account unlocks, software installation guidance, connectivity troubleshooting, basic how-do-I questions - follows structured scripts that AI handles reliably. Tier-2 escalations - hardware failure diagnosis, network performance issues, software conflicts - are partially automatable. Tier-3 escalations - novel system failures, security incidents, complex multi-system interactions - retain the strongest human requirement.
The 7.5/10 score reflects this gradient: the bulk of support volume sits in tiers 1 and 2, where AI is now competent. The high-complexity minority that remains fully human is a smaller share of total positions than the score alone suggests.
What is already being automated
| IT support task | Automation tool type | Status | AI risk |
|---|---|---|---|
| Password resets and account unlocks | Self-service AI portal | Fully automated | 9.5/10 |
| Software install / update guidance | AI chatbot (Zendesk, ServiceNow) | Largely automated | 8.5/10 |
| Basic network connectivity troubleshooting | AI diagnostic + script | Largely automated | 8.0/10 |
| Ticket routing and prioritisation | AI classification | Fully automated | 9.0/10 |
| Hardware fault diagnosis | AI-assisted, human-led | Partially automated | 6.0/10 |
| Security incident first response | SIEM + AI alert, human decision | Augmented | 5.5/10 |
| On-site hardware repair and cabling | Physical presence required | Not automated | 2.0/10 |
Why network and security technicians are safer
The 7.5/10 average across ISCO 35 conceals an important bifurcation. Network technicians who manage physical infrastructure - running cable, replacing switches, diagnosing physical-layer failures - face a much lower AI risk because their work requires on-site presence and hands-on diagnostic judgment. A fibre splice, a rack installation, a physical server failure diagnosis: these are not automatable by software regardless of how capable the AI is.
Cybersecurity operations roles are more nuanced. AI tools including CrowdStrike Falcon, Microsoft Sentinel, and Darktrace are increasingly capable at detecting anomalies, flagging events, and generating alert summaries. But the decision loop for incident response - is this a genuine breach? what is the scope? what should we isolate first? when do we call the CISO? - retains strong human accountability requirements. Organisations are not yet willing to let an AI make autonomous containment decisions in a live security incident.
The safest IT roles from AI
| Role type | Why lower risk | AI score |
|---|---|---|
| On-site hardware technician | Physical presence, repair, hands-on diagnosis | 2.5/10 |
| Network infrastructure tech | Physical cabling, equipment installation, site-specific | 3.5/10 |
| Cybersecurity operations | Human judgment in incident response, accountability | 5.5/10 |
| Tier-2 escalation specialist | Complex multi-system troubleshooting, some AI assist | 6.5/10 |
| Tier-1 helpdesk agent | Script-following - core AI use case | 9.0/10 |
What this means for you
If you work in tier-1 helpdesk, the honest picture is that your core role is the primary use case for enterprise AI support tools. Companies are not replacing helpdesk agents all at once - they are deploying AI-first routing that handles the bulk of tickets automatically and reserves human agents for escalations. The headcount implication: fewer tier-1 agents, more pressure on tier-2, and growing demand for the people who configure and maintain the AI support tools themselves.
The career path that holds value through this transition is upward: from ticket-handler to system-configurer, from troubleshooter to incident analyst, from reactive support to proactive monitoring. An IT support professional who understands how AI helpdesk tools work, can write and maintain automation scripts, and can diagnose the failures AI cannot handle is substantially more resilient than one who cannot. The path to job security in this occupation runs directly through developing the skills to configure the systems that are competing with you.
Explore IT exposure by country
See how AI affects technology workers across 206 countries.
Open the Explore Tool →Was this analysis useful?