IT · Obsolescence report
Will AI replace IT jobs?
AI will not replace IT as a field, but it is already replacing a lot of routine IT work: first-line support tickets, boilerplate code, standard configs and test scripts. Roles built around troubleshooting real systems, security and architecture are holding up best.
Doomed
Across the 11 it roles below, AI exposure averages 56% (Doomed). The most exposed is qa tester at 66%; the safest is network engineer at 45%.
IT and computer science jobs sit right in the path of AI because so much of the work is text: code, configs, tickets and documentation. That makes the routine parts fast to automate, and it also means IT people are among the first to benefit from AI tools.
The jobs below are ranked by how much of their typical week AI can already do. Coding-heavy roles score highest; roles that own infrastructure, security and incident response score lower.
11 IT jobs ranked by AI exposure
| # | Job | AI exposure | Verdict |
|---|---|---|---|
| 1 | QA Tester | 66% | Doomed |
| 2 | Data Engineer | 59% | Doomed |
| 3 | Programmer | 58% | Doomed |
| 4 | Web Developer | 57% | Doomed |
| 5 | Software Engineer | 57% | Doomed |
| 6 | Database Administrator | 55% | Doomed |
| 7 | Cybersecurity Analyst | 55% | Doomed |
| 8 | Mobile App Developer | 54% | Endangered |
| 9 | IT Support Specialist | 53% | Endangered |
| 10 | DevOps Engineer | 52% | Endangered |
| 11 | Network Engineer | 45% | Endangered |
Most at risk
- QA Testers (66%): writing and maintaining scripted regression tests is already automating.
- Data Engineers (59%): writing standard SQL, transformations, and pipeline boilerplate is already automating.
- Programmers (58%): writing boilerplate, CRUD endpoints, and glue code is already automating.
Safest
- Network Engineers (45%): designing networks that balance cost, security, and resilience stays human.
- DevOps Engineers (52%): owning production reliability when everything is on fire stays human.
- IT Support Specialists (53%): physically fixing, swapping, and setting up hardware stays human.
Where it work goes
The average split of a working week across these roles, and how exposed each kind of task is to AI:
| Task type | Share of time | AI exposure |
|---|---|---|
| Writing code | 31% | High |
| Routine analysis & reporting | 9% | High |
| Judgment & strategy | 19% | Low |
| Research & synthesis | 7% | High |
| Data entry & processing | 5% | High |
| Scheduling & admin | 6% | High |
| Writing & drafting | 4% | High |
| Scripted customer communication | 4% | High |
| Trust & relationships | 6% | Low |
| Licensed accountability | 5% | Low |
How to stay ahead
Move up the stack. Let AI write the scripts, answer the routine tickets and draft the docs, and build skills in architecture, security, cloud cost and reliability, where mistakes are expensive and someone has to own them.
Tools for the most exposed work: AI Coding, Automation & Agents, Productivity.
Get your personal score
Your job is not the average. Answer four questions and get your own exposure score and survival plan.
Will AI take my job? →Questions the Humans Keep Asking
Will AI replace IT jobs?
AI will not replace IT as a field, but it is already replacing a lot of routine IT work: first-line support tickets, boilerplate code, standard configs and test scripts. Roles built around troubleshooting real systems, security and architecture are holding up best. Across the 11 it roles below, AI exposure averages 56% (Doomed). The most exposed is qa tester at 66%; the safest is network engineer at 45%.
Which IT jobs are most at risk from AI?
By task exposure: qa testers (66%), data engineers (59%), programmers (58%).
Which IT jobs are safest from AI?
By task exposure: network engineers (45%), devops engineers (52%), it support specialists (53%).
Which IT jobs are most at risk from AI?
Entry-level support, manual QA testing and routine coding are the most exposed, because the work is repetitive and well specified.
Is computer science still worth studying?
Yes, but the entry-level coding job is changing. Graduates who understand systems, security and how to direct AI tools will have a clear advantage over those who only write routine code.